TYPEFACE PERSONALITY TRAITS

AND THEIR DESIGN CHARACTERISTICS

Ying Li

A Thesis

In

The Department

of

Computer Science and Software Engineering

Presented in Partial Fulfillment of the Requirements

For the Degree of Master of Computer Science at

Concordia University

Montreal, Quebec, Canada

November 2009

© Ying Li, 2009

 Library and Archives Bibliothèque et
1*1 Canada Archives Canada

Direction du
Published Heritage
Branch Patrimoine de l'édition

395 Wellington Street 395, rue Wellington

Ottawa ON K1A 0N4 Ottawa ON K1A 0N4
Canada Canada

Your file Votre référence

ISBN: 978-0-494-71016-6
Our file Notre référence
ISBN: 978-0-494-7 1 0 1 6-6

NOTICE: AVIS:

The author has granted a non- L'auteur a accordé une licence non exclusive
exclusive license allowing Library and permettant à la Bibliothèque et Archives
Archives Canada to reproduce, Canada de reproduire, publier, archiver,
publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public
communicate to the public by par télécommunication ou par l'Internet, prêter,
telecommunication or on the Internet, distribuer et vendre des thèses partout dans le
loan, distribute and sell theses monde, à des fins commerciales ou autres, sur
worldwide, for commercial or non- support microforme, papier, électronique et/ou
commercial purposes, in microform, autres formats.
paper, electronic and/or any other
formats.

The author retains copyright L'auteur conserve la propriété du droit d'auteur

ownership and moral rights in this et des droits moraux qui protège cette thèse. Ni
thesis. Neither the thesis nor la thèse ni des extraits substantiels de celle-ci
substantial extracts from it may be ne doivent être imprimés ou autrement
printed or otherwise reproduced reproduits sans son autorisation.
without the author's permission.

In compliance with the Canadian Conformément à la loi canadienne sur la

Privacy Act some supporting forms protection de la vie privée, quelques
may have been removed from this formulaires secondaires ont été enlevés de
thesis. cette thèse.

While these forms may be included Bien que ces formulaires aient inclus dans
in the document page count, their la pagination, il n'y aura aucun contenu
removal does not represent any loss manquant.
of content from the thesis.

1+1

Canada
 ABSTRACT

Typeface Personality Traits and Their Design Characteristics

Ying Li

Typography is a critical tool in visual communication. Selecting the appropriate

typeface to express and communicate a message is very important. Since most studies
on typeface design concentrate on typeface legibility and readability, this thesis
focuses on the visual expression of typefaces and their design characteristics. The
relationship between typefaces and their personality traits including legible, cheerful,
fearful, creative, attractive, formal, sloppy, relaxed, friendly and confident are
investigated.
A font survey about twenty four typefaces and ten personality traits is conducted and a
series of statistical analyses are performed to discover the correlation between
typefaces and their personality traits. As a result, the number of studied typefaces is
reduced from twenty four to fifteen and these fifteen typefaces are categorized into
four groups according to their personality traits and typographical features.
Typeface design characteristics, such as x-height proportion, ascender and descender
proportion, font weight, stroke design, counter design as well as character space of
these fifteen typefaces are studied in depth. Typeface design characteristics of four
different groups are summarized.
In addition, the aesthetic design characteristics of studied typefaces are analyzed. The
appropriate uses of each of the four groups are discussed.

Ill
 Acknowledgements

First I would like to express my sincere gratitude to my supervisor, Dr. Ching Y. Suen,

for his insightful advice and invaluable encouragement, which have helped me

through my studies at Concordia.

I would further like to express my gratitude to all of my colleagues and friends in

CENPARMI. Thanks to Chunlei He, Yan Zhang, Dr. Wumo Pan, Xiaoxiao Niu,

Nicola Nobile, Shira Katz, Guiling Guo, etc., who helped me in one way or another.

I would also thank Ms. Marleah Bloom for her help on collecting survey data, as well

as her excellent editing and proofreading work, which have contributed to

dramatically improve the expression of this thesis.

Finally, I would like to thank my parents. Without their continuous support and

encouragement, I would not have been able to complete this thesis.

CONTENTS
List of Figures viii
List of Tables x

Chapter 1 Introduction 1
1.1 Motivation and Objectives !
1.2 Typeface Terminology 3
1.2.1 Anatomy ofTypeface 3
1.2.2 Typeface Classifications 6
[Link] Classification Based on Historical Development 6
[Link] Classification Based on Visual Appearance 7
1.2.3 Typeface Selection and Usage 8
[Link] Type Size Measurement · 8
[Link] Legibility and Readability 9
[Link] Typeface Weight 9
[Link] Character Spacing 9
1.3 Thesis Outline 10
Chapter 2 Typeface Personality Survey 12
2.1 Literature Review ofTypeface Personality Studies 12
2.2 Proposed Typeface Personality Study Method 13
2.2.1 Studied Typefaces 14
2.2.2 Typeface Personality in Research I5
2.2.3 Rating Scale 1&
2.2.4 Participants 16
2.2.5 Materials and Procedure 17
2.2.6 Data Collection Methods 19
Chapter 3 Analysis of Survey Results 20
3.1 Univariate Analysis 21
3.1.1 Distributions of Typeface Rating Scores 21
3.1.2 Measures of Central Tendency and Spread 23
3.2 Correlation Analysis 26
3.3 Factor Analysis ··· 30
3.3.1 Kaiser-Meyer-Olkin Measure of Sampling Adequacy 31
3.3.2 Deriving Principal Component 31
3.3.3 Varimax Rotation 33
3.3.4 Interpretation of Factors 34
3.4 Multidimensional Scaling (MDS) Analysis 36
3.5 Demographic Differences 37
3.6 Discussion 39
Chapter 4 Typeface Characteristics Analysis 43
4.1 Typographical Design Characteristics 43
4.1.1 Legibility 44
4.1.2 x-height Proportion 45
4.1.3 Ascender Proportion 48
4.1.4 Descender Proportion 50
4.1.5 Weight Detection 52
4.1.6 Serif and Sans Serif 54
[Link] Serif 56
[Link] Sans Serif 57
[Link] Display 57
[Link] Legibility between Serif and Sans Serif 58
4.1.7 Character Stroke Contrast Design 59
4.1.8 Character Width and Height Design 61
4.1.9 Stem and Cap Height Design 64
4.1.10 Character Stroke Design 67
[Link] Stroke Length 67
[Link] Stroke Form 69
[Link] Stroke Joining Part Treatment 70
4.1.11 Counter Design 71
4.1.12 Character Space 72
vi
4.2 Aesthetic Design Characteristics 73
4.2.1 Elaborateness 74
[Link] Ornament vs. Briefness 74
[Link] Depth vs. Flatness 75
[Link] Special Use and Common Use 76
4.2.2 Naturalness 76
[Link] Organic vs. Geometric 76
[Link] Printed vs. Handwritten Appearance 77
4.2.3 Harmony 78
[Link] Symmetry vs. Asymmetry 78
[Link] Balanced vs. Unbalanced 79
4.3 Summary of Typeface Characteristics 79
4.3.1 Typographical Characteristics 79
[Link] Typographical Characteristics of Four Groups 80
[Link] Typographical Characteristics of Groups' Representative Typefaces . 82
4.3.2 Aesthetic Characteristics 87
4.3.3 Appropriate Uses 89
Chapter 5 Conclusion 92
5.1 Summary "2
5.2 Future Work 93
References 9->
Appendix A Order of Typefaces in the Survey 97
Appendix B Sample of the Font Survey 98
Appendix C Name Abbreviation of Typeface 128
Appendix D Name Abbreviation of Personality Trait 129
Appendix E Pearson's Correlation Coefficients 130
Appendix F Rotated Component Matrix 141
Appendix G Fifteen Examined Typefaces 151

vii
List of Figures
Figure 1 Anatomy of typeface 4
Figure 2 More anatomy of typeface 5
Figure 3 Examples of typeface classification based on
the historical development 7
Figure 4 Examples of Serif and Sans Serif typefaces 8
Figure 5 Example ofmono-spaced typeface 10
Figure 6 Example of proportional typeface 10
Figure 7 Twenty four typefaces used in the survey 14
Figure 8 Sample of the alphabets and text displayed in the
font survey. This sample shows the typeface Poor Richard 18
Figure 9 Normal distribution of rating scores of typeface Harrington
related to personality trait "Cheerful" 22
Figure 1 0 Slightly skewed distribution of rating scores of
typeface Rockwell related to personality trait "Relaxed" 22
Figure 1 1 Slightly skewed distribution of rating scores of
typeface Times New Roman related to personality trait "Confident" 23
Figure 12 Pearson correlation coefficient 27
Figure 1 3 MDS analysis of fifteen typefaces related to the
personality trait "Friendly" 37
Figure 14 Histogram of typeface "Garamond" concerning
ten personality traits in male and female groups 38
Figure 15 Four typographical lines from vertical projection profiles 46
Figure 16 x-height proportion of fifteen typefaces in 36 point 47
Figure 17 Ascender/height of fifteen typefaces in 36 point 49
Figure 18 Descender/height of fifteen typefaces in 36 point 51
Figure 19 Horizontal projection profiles 52
Figure 20 Weight of fifteen typefaces in 36 point 54
Figure 21 Upper end of Capital "C", implied spur, bracketed spur and
viii
 Slab Serif (from left to right: Centaur, Times New Roman and
Rockwell) 57
Figure 22 Ends of Capital C of Sans Serif (from left to right: Arial,
Helvetica and Kabel) 57
Figure 23 Capital Cs (from left to right: Harry Potter, Snap ITC,
Jokerman, Chiller and Harrington) 58
Figure 24 Examples of Capital U (from left to right: Centaur,
Times New Roman, Arial, Rockwell, Bauhaus 93 and
Harry Porter in 36pt) 59
Figure 25 Stem height and Cap height of Capital "Y" 64
Figure 26 Ornaments of typeface "Jokerman" 75
Figure 27 Curvaceous strokes of typeface "Harrington" 75
Figure 28 Curvaceous strokes of typeface "Belwe Lt BT" 75
Figure 29 Geometric and even strokes of typeface "Bauhaus 93" 77
Figure 30 Curly Serifs of typeface "Harrington" 77
Figure 31 Shivery chilled strokes of typeface "Chiller" 78

ix
List of Tables

Table 1 Twenty four typefaces used in the study 15

Table 2 Ten adjectives used to assess font personalities 16
Table 3 Five-point modified likert scale 16
Table 4 Mean values ofrating scores of twenty four typefaces
related to ten personality traits 24
Table 5 Five typefaces that were the most associated with
each of the ten personality traits and their means 26
Table 6 Correlation matrix of twenty four typefaces related to
personality trait "Legible" 29
Table 7 Rotated component matrix of fifteen typefaces related to the
personality trait "Cheerful" 34
Table 8 Four groups and their corresponding typefaces 36
Table 9 Comparisons of experimental method between Shaikh (2006)
and our studies 41
Table 1 0 Comparisons of study results between Shaikh (2006) and our studies 42
Table 11 Fifteen studied typefaces in four groups 43
Table 1 2 Mean values of rating scores of personality trait "Legible"
for fifteen typefaces within their corresponding groups 44
Table 13 Legibility comparison of four groups 45
Table 14 x-height proportions of fifteen typefaces in 36 point 47
Table 15 Ascender proportions of fifteen typefaces in 36 point 49
Table 16 Descender proportions of fifteen typefaces in 36 point 51
Table 17 Font weight of fifteen typefaces in 36 point 53
Table 18 Fifteen typefaces classified based on
typographical features, classification refers to [10] 55
Table 19 Capital "C" of fifteen typefaces in 36 point 56
Table 20 Ratios between the two stems of U for fifteen typefaces 60
Table 21 Capital "O" in fifteen typefaces in 36 point 61

X
Table 22 Ratio between width and height of O for all fifteen typefaces 63
Table 23 Capital "Y" for all fifteen typefaces in 36 point 65
Table 24 Ratios between stem and Cap height ofY of fifteen typefaces 65
Table 25 Capital "E" of fifteen typefaces in 36 point 67
Table 26 Capital "D" of fifteen typefaces in 36 point 69
Table 27 Capital "M" of fifteen typefaces in 36 point 70
Table 28 Lowercase "b" of fifteen typefaces in 36 point 71
Table 29 Left side bearings (LSB) and right side bearings (RSB)
of characters "H", "0", "h" and "o" for fifteen typefaces 73
Table 30 Mean values of four main typographical characteristics of
Our four study groups 80
Table 3 1 Five typefaces that were the most associated with
personality traits "Legible", "Formal" and "Confident" 82
Table 32 Means of four main typographical characteristics of
five typefaces in the Directness group 83
Table 33 Means of four main typographical characteristics of
typefaces in the Gentleness group 84
Table 34 Five typefaces that were the most associated with
personality traits "Cheerful", "Attractive", "Creative" and "Relaxed" .... 84
Table 35 Means of four main typographical characteristics of typefaces
in the Cheerfulness group 85
Table 36 Five typefaces that were the most associated with
personality traits "Fearful", "Creative" and "Sloppy" 86
Table 37 Means of four main typographical characteristics of typefaces
in the Fearfulness group 86
Table 3 8 Different levels of aesthetic characteristics of four study groups 88

xi
Chapter 1 Introduction

In this chapter, the motivation and objectives of this thesis are introduced. We also

review typography terminologies and some typeface design characteristics. Lastly, we

present the structure of this thesis.

1.1 Motivation and Objectives

In this thesis, we focus on the visual expression of typefaces and their design

characteristics. The relationship between typefaces and their personas is investigated.

By using statistical analyses on data collected from participants who filled out a

survey, the correlation between fonts and personas is explored. Fonts used within this

study are grouped according to their personas and typical characteristics of typefaces

in these groups are examined in detail.

Typography is a critical tool in visual communication, because typeface can evoke

human emotions. Due to different styles and a variety of proportions, weights, heights,

etc., each typeface has its own aesthetic and expressive qualities, as evidenced by the

visual attributes of its letterforms [I]. Some fonts can reinforce a chosen message,

whereas others can detract from an intended meaning and have adverse effects.

Therefore, selecting the appropriate typeface to express and communicate a message

is very important.

?
Each typeface has its own individual identity. In a BBC audio program on February

11th, 2005, Ian Peacock [2] explores how the fonts we choose are sending secret

subliminal messages about who we are. He argues that the fonts we use to dress our

words are as much of a fashion statement as the clothes we wear. Within the program,

fonts were also depicted as being feminine or masculine, as well as possessing other

traits. Feminine fonts, for example, were described as fine, serif, sleek, and elegant,

while masculine fonts were characterized as being blocky and bold.

Most research on fonts is related to legibility and readability. There are only a few

studies on typefaces and their potential personas. In our study, we examine whether

specific typefaces are perceived to have particular personality traits. First, we

established ten different personas for twenty four typefaces. A survey was then

created and administrated to individuals who voluntarily participated in the study.

This survey was created to help determine whether or not participants think that

the twenty four chosen fonts are associated with ten tangible personality traits, and to

what degree fonts can convey these traits.

After obtaining sufficient data, our next step was to analyze how particular typefaces

are associated with certain personality traits. In order to measure the relationship

between typefaces and personas quantitatively, we used standard statistical methods

to evaluate the relationship between studied typefaces and personality traits.

The relationship between typefaces and personality traits are thus examined. In this

thesis, typeface design characteristics, such as x-height proportion, ascender and

descender proportion, font weight as well as stroke design and so on are studied

2
further. We also analyze the aesthetic design characteristics of studied typefaces.

1.2 Typeface Terminology

In this section, we review typeface terminologies, including typeface and font,

anatomy of typeface, typeface classification and typeface selection and usage.

1.2.1 Anatomy of Typeface

Letterforms are sets of letters, numbers and other symbols. A typeface is a set of one

or more fonts, in one or more sizes. It is designed with stylistic unity as each typeface

is comprised of a coordinated set of glyphs. Arial and Times New Roman are two

examples of typefaces. A font is a particular example of a typeface, with a particular

size, weight and angle. For example, 8-point Arial, 10-point Arial and 10-point Arial

Italic are three different fonts but are all members of the Arial typeface. A glyph is a

single representation of a typographic character in a typeface.

In Figures 1 and 2 below, the terms such as baseline and x-height, are included to help

understand and describe the typeface anatomy.

3
 ascender
CAP-HEIGHT --
serifs

X-HEICHT

BASELINE

descender

Figure 1 Anatomy of typeface

x- height: the basic height of the lowercase letter x. The x-height can vary greatly

from typeface to typeface at the same point size.

Baseline: the line on which all letters rest.

Cap-height: the distance from baseline to cap line of an alphabet, this is the

approximate height of the uppercase letters.

Ascender: the part of some lowercase letters (such as b, h or d) which ascends above

the x-height.

Descender: the part of some lowercase letters (such as y, ? or q) that descends below
the baseline.

Serif: a stroke added to the beginning or end of one of the main strokes of a letter.

Contrast: the degree of difference between the thick and the thin strokes in a

letterform.

4
 U¡M<X
stem

!«ml

,
syoc
terminal '
_/"
/ vertex
^J
V1
* axis
! /
l'Olii !UT

Figure 2 More anatomy of typeface

Axis: the axis of a letter means the axis of the stroke, which in turn reveals the axis of

the pen or other tool used to make the letter.

Bowl: the round or elliptical forms which are the basic body shape of letters such as C,

G, O in the upper case, and b, c, e, ?, ? in the lower case. It is also called eye.

Stem: a main stroke that is more or less straight, not part of a bowl. The letter o, for

example, has no stem; the letter 1 consists of stem and serif alone.

Counter: the white space enclosed by a letterform, whether wholly enclosed, as in d or

o, or partially, as in c or m.

Terminal: a curved stroke, which is usually apparent on the tail or stem of some letters

(such as j, y, r and a). It is not a serif.

Apex: the uppermost point of a character where the vertical strokes meet.

Vertex: the bottom of a letter where two straight strokes or stems join and create an

angle, such as in V, Y and W.

Crossbar: a horizontal stroke or arm that connect two stems (as in H or A).

5
1.2.2 Typeface Classifications

There are many scales to classify different typefaces. We introduce two typical scales,

which include classification based on historical development and classification based

on visual appearance.

[Link] Classification Based on Historical Development

According to the French typographic historian Maximilian Vox, typefaces can be

classified within six main groups based on specific historical periods. These groups

include pre-Venetian (before 1400), Venetian (1400-1500), Garalde which is also

called Old Roman or Old Style (1600), Transitional (1700), Didone (1700-1900) and

Display (1900-present) [3]. Some typeface examples classified based on historical

development are shown in Figure 3.

6
 graphique
Venetian Centaur

graphique
Garalde Garamond

graphique
Transitional Baskerville Old Face

graphique
Didone Rockwell

graphique
Display Comic Sans

Figure 3 Examples of typeface classification based on the historical

development

[Link] Classification Based on Visual Appearance

Typefaces can be classified into three categories based on their visual appearance.
These categories are Sans Serif, Serif and decorative typefaces.

Serif is the typeface with small features at the end of strokes within letters. The

typefaces without serifs are considered Sans Serif (from French sans, meaning

without) (Figure 4).

7
 APsIlIA Fit
Figure 4 Examples of Serif and Sans Serif typefaces
(from left to right: Serif and Sans Serif)

A decorative typeface differs as it involves a particular use of typeface. These

typefaces may be used for headlines and not appropriate for text documents. The best

appearance of decorative typefaces are at large display sizes, typically 36 points or

larger.

1.2.3 Typeface Selection and Usage

There are many factors that influence the selection of a typeface. Type size
measurement, legibility and readability, weight and space are all influential factors

that are presented in this section.

[Link] Type Size Measurement

Standard type face sizes range from 4 up to 120 points, where a point is the smallest

typographical unit of measurement. Each point measures 0.0138 of an inch, which is

equivalent to 1/72 of an inch. There are approximately 72 points (0.9936 inch) to one

inch. A pica is 12 points (0.1660 inch). There are approximately 6 picas (6.0230) to
one inch.
[Link] Legibility and Readability

Legibility and readability are two important aspects of a typeface. Legibility means

the quality of being easy to read, and it is the term used when discussing the clarity of

single characters. Readability is the term which describes the quality of visual comfort,

an important requirement in the comprehension of long stretches of text [4].

[Link] Typeface Weight

The weight of a typeface is reflected in the typeface's design or style. It is the visual

lightness or darkness of form. A typeface family may offer a full range of weights

from light (or thin) to extra bold.

[Link] Character Spacing

The interrelationship between the white space and the text in a typeface is also an

important aspect. Roughly, there are two different scales of character spacing,

mono-spaced typeface (Figure 5) and proportional typeface (Figure 6). Mono-spaced

typeface means each character fits into the same character width, while proportional

typeface means that each character width is different in order to accommodate the

particular width of each character.

9
 mo ? os pa cfeldl
|t|y|p|e|f|a ce
Figure 5 Example of mono-spaced typeface

rtional
PlTO0 ,
tkpeface
Figure 6 Example of proportional typeface

1.3 Thesis Outline

This thesis consists of four additional chapters.

In the next chapter, we present a description of the study, which investigates typefaces

and their personalities. This description includes an overview of the font survey that

was used to investigate the relationship between twenty four typefaces and ten

personalities. Research methodology, including a description of participants, materials,

data collection and procedure of font survey are also outlined in this chapter.

Chapter 3 focuses on the statistical analysis. We used SPSS (version 17.0) to analyze

the data collected from the font survey, including various methods of analysis such as

Correlation, Factor Analysis, Multidimensional Scaling and one way Analysis of

Variance, etc.
In Chapter 4, we examine the design characteristics of fifteen typefaces from

typographical design and aesthetics. The measurement and analysis on typical design

factors of these typefaces are presented.

Chapter 5 provides our conclusions and suggestions on topics for future exploration
based on our research results.

11
Chapter 2 Typeface Personality Survey

In order to investigate whether or not viewers associate particular typefaces with

emotional qualities, we developed and administrated a font survey on the relationship

between twenty four fonts and ten personalities. This chapter begins with a literature

review of different studies on typeface personalities. A description of the font survey

and the methodology used within this study are then presented.

2.1 Literature Review of Typeface Personality

Studies

Most research on typefaces is related to font legibility and readability. There are,

however, a few studies on personalities that fonts may have, personalities that convey

messages beyond what is expressed within the text. In the area of marketing and

consumer psychology, typeface personality has been studied for a long time. The

earliest study is by Proffenberger and Franken [5], who identified five atmosphere

qualities for twenty nine typefaces. These qualities include cheapness, dignity,

economy, luxury, and strength. Subsequently, Spencer [6] mentioned in his book that

typefaces can be grouped under three headings of atmosphere value:

luxury/refinement, economy/precision and strength. Some researchers assigned

specific personas to specific typefaces. Kostelnick, Roberts and Dragga [7] depicted

12
Times New Roman as "booklish and traditional"; Bodoni as "dramatic and

sophisticated" and Goudy as "corpulent and jolly". Shunshan and Wright [8]

described Garamond as "graceful, refined and confident" and Century Schoolbook as

"serious yet friendly". Some typographers also have perceived that particular

typefaces are imbued with cultural and national characteristics. Laliberte [9] attributed

several typefaces to represent several countries, Fraktur for Germany, Garamond for

France, Bodoni for Italy, and Caslon for England.

There are however, discrepancies within these past studies on the topic of typefaces

and their associated personalities. The personalities identified by the above stated

researchers are not consistent. This may be due to the difference in participants based

on gender, age or other demographic factors. Consequently there are discrepancies

within findings from past studies.

2.2 Proposed Typeface Personality Study Method

In our study, a survey with twenty four different fonts in two sizes and ten

personalities was developed to help determine whether or not viewers think that the

chosen fonts are associated with tangible personality traits, and to what degree these

fonts convey these traits.

13
 2.2.1 Studied Typefaces

Cooper Black Berlin Sam FB Bernard MT COIldeiK ed

Garamond Belwe Lt BT Haylrill Wt/ Poffer Centaur
PoorRicliarcl CFo&efman Times New Roman
Arial Broadway Kino MT Impact Chiller
Helvetica ßauhou/93 Kabel Op Rockwell
Sitap ITC Harrington Fo otüghtiMT Light
Figure 7 Twenty four typefaces used in the survey

Twenty four different typefaces were chosen as test typefaces (Figure 7). We selected
these twenty four typefaces to represent a wide range of physical characteristics from

Serif and Sans Serif to display typefaces (see Table 1 for a complete listing,
classification refers to [10]). Each typeface exhibits variations in typeface design from

x-height, ascender, descender and stroke weight, etc. Also, these twenty four
typefaces are widely used in different applications. Some of them are standard and

most frequently used in books and newspapers, such as Times New Roman and Arial.

Others, such as Cooper Black, Impact and Broadway, are popular for advertising.

14
 Serif
Centaur
Garamond
Times New Roman
Sans Serif
Arial
Helvetica
Berlin Sans FB

Slab Serif
Rockwell
Playbill
Display Serif
Cooper Black
10 Bernard MT Condensed
11 Onyx
12 Footlight Light
13 Poor Richard
14 Belwe Lt BT
Display Sans Serif
15 Impact
16 Kabel
Display
17 Bauhaus 93
18 Broadway
19 Harrington
20 Kino MT
21 Snap ITC
22 Jokerman
23 Chiller
24 Harry Potter
Table 1 Twenty four typefaces used in the study

2.2.2 Typeface Personality in Research

We selected ten typeface personality traits (Table 2) based on previous studies. These

studies have frequently referred to such adjectives to describe typefaces within the
literature.

15
 Cheerful
Fearful
Legible
Attractive
Creative
Formal
Sloppy
Relaxed
Friendly
Confident

Table 2 Ten adjectives used to assess font personalities

2.2.3 Rating Scale

Not at all | Slightly Moderately Highly Extremely

Table 3 Five-point modified likert scale

We used a modified five point Likert Scale with the categories as shown in Table 3.

The scale was used to reflect a range of different responses from participants to the

twenty four typefaces.

2.2.4 Participants

The participants were Concordia University students and staff, as well as others who

were interested in this topic. The respondents were recruited through e-mails and

posters in Concordia University.

A total of 75 participants completed the survey, 37 females and 38 males.

Approximately 58.7% of participants were between 20-29 years of age, and 22.7%
16
between 30-39 years. Only one participant was younger than 20 years and the

remaining 17.3% participants were older than 40 years.

Approximately 40% of respondents reported having a bachelor degree, 42.7% a

master's degree and 10.7% a doctorate. The education backgrounds of the remaining

6.6% participants include High School, Technical School and Junior College.

2.2.5 Materials and Procedure

For each typeface, the complete alphabet in 22 points was displayed in an image that

included capitals, lower cases and numerals. Two pangrams, "The quick brown fox

jumps over the lazy dog" and "Please complete the survey to your comfort level"

were also displayed in 16 points in another corresponding image. Figure 8 illustrates a

sample of the display participants were given for each of the twenty four typefaces.

The text samples were converted to binary images at 200*200 dpi resolution.

17
 ABCDEFGHIJKLMNOPQRSTUVWXYZ
a L· e cl e ? g 1? i j k 1 m ? ? ? g ï s t u ? w ? ij ?
012345Ô789

1lie cjuick trown tox jumps oveï

tlie lazu clo£. Please complete
ite surveg io vjom" comiori level.

Figure 8 Sample of the alphabets and text displayed in the font survey. This
sample shows the typeface Poor Richard

The twenty four typefaces were randomly distributed throughout the survey to avoid

any effects due to order. The order of twenty four typefaces displayed in the font

survey is provided in Appendix A. The two images were presented at the top of each

page, followed by the rating scale.

The survey was provided as printed and online forms, with 27 questions. 24 questions

addressed the twenty four fonts and ten personality traits, and 3 questions inquired

about demographic information, including age, gender and education background.

The survey took approximately 30-35 minutes to complete. Participation in the survey

was voluntary and participants are able to discontinue the survey at any time, without

consequence, in accordance with Concordia University's Office of Research Policies.

Participants were provided written instructions at the beginning of the survey. They

were asked to visually examine the computer or paper displays of the twenty four

typefaces and rate them on ten personality traits, indicating how well the typeface

suited each personality trait. A copy of the survey is provided in Appendix B.

18
2.2.6 Data Collection Methods

We used the online survey tool - freeonlinesurvey [1 1] to create the survey form. Two

approaches were used to conduct the survey, the online version and the paper version.

This online survey was administered by the survey tool, including data collection,

data storage, etc. Participants of the online survey were recruited via inviting emails.

They followed the survey link and completed the survey online.

The data for the paper survey were inputted by us and data storage was also

administered by the survey tool. Participants were paid for completing the paper

survey. We distributed the hard copy to prospective participants, and they returned the

completed survey to get the monetary reward.

19
 Chapter 3 Analysis of Survey Results

In this chapter, we describe the survey results, which were analyzed by using data

with statistical software SPSS (version 17.0).

First, we performed univariate analysis on survey data and calculated the central

tendency and dispersion of the rating scores of each typeface's personality traits.

Second, we performed correlation analysis on the survey data to determine the

relationship between each of the twenty four typefaces related to every personality

trait. We could then decide whether correlations exist between any of our studied

typefaces.

Third, because the correlation analysis of twenty four typefaces and ten personality

traits produced too numerous and detailed information for analysis and presentation,

we removed several typefaces whose relationship strength were not statistically

significant based on the correlation analysis results. We then conducted factor

analysis on the remaining typefaces to group them into smaller sets and identify

common underlying factors.

Fourth, we applied multidimensional scaling to survey data for the remaining

typefaces related to every personality trait to validate the results of the factor analysis.

Lastly, we analyzed the survey's demographic data to examine its potential influence

on participants' responses.

The following sections discuss these analyses in greater detail.

20
3.1 Univariate Analysis

In order to explore the characteristics of individual variables from our survey data and

to prepare for further statistical procedures, we first applied univariate analysis on the

rating scores for each typeface of the survey data.

3.1.1 Distributions of Typeface Rating Scores

We examined the histogram of rating scores of each typeface to explore the

distributions of rating scores for each typeface related to each personality trait.

Through the analysis we found that, the histograms of rating scores exhibited two

common shaped distributions: normal and slightly skewed. These two types of

distributions are illustrated in Figure 9 (typeface Harrington related to personality trait

"Cheerful"), Figure 10 (typeface Rockwell related to personality trait "Relaxed") and

Figure 1 1 (typeface Times New Roman related to personality trait "Confident").

21
 C
3
O
?

Hr Ch

Figure 9 Normal distribution of rating scores of typeface Harrington related to

personality trait "Cheerful"

30.(H

C
3
O
?

.00 1.00 2.00 3.00 4.00 5.00 6.00

Rw Rx

Figure 10 Slightly skewed distribution of rating scores of typeface Rockwell

related to personality trait "Relaxed"

22
 30.(H

20-OH
C
3
O
?

.00 1.00 2.00 3.00 4.00 5.00 6.00

TNR Cn

Figure 11 Slightly skewed distribution of rating scores of typeface Times New

Roman related to personality trait "Confident"

3.1.2 Measures of Central Tendency and Spread

We examined the mean values, minimum values, maximum values and standard

deviations of rating scores of each typeface based on each personality trait. Table 4 is

the mean values of rating scores of twenty focr typefaces re'aiec ?? ?at

traits. We summarized the five typefaces that were the most associated with each of

the ten personality traits and their mean values in Table 5.

The abbreviations used for typefaces are shown in Appendix C and in the next page

for easy reference, and the abbreviations used for personality traits is shown in

Appendix D, similarly hereinafter.

23
li.

ß
O
vi
U
<a
a.
a

?«

?
a

U
S

?«
?
U

tu

ü ?
Name abbreviation of typeface

Typeface Name Abbreviation

Cooper Black CB
Berlin Sans FB BSF

Bernard MT Condensed BMC

Garamond Ga
Belwe Lt BT BLB

Playbill Pb

Harry Potter HP
Centaur Cr
Poor Richard PR
Jokerman Jm
Times New Roman TNR
Arial Al

Broadway Bw

Kino MT KM

Impact Ip
Chiller Cl
Helvetica Ht
Bauhaus 93 Bh93
Kabel Kb

Onyx Ox
Rockwell Rw

Snap ITC SITC

Harrington Hr

Footlight MT Light FL

25
 Personality Typeface
Jm SITC Hr BLB Ga
Cheerful
3.4533 3.2933 3.0133 2.6533 2.6000

Hp Pb Cl BMC KM
Fearful
2.9067 2.4800 2.3867 2.1067 2.0933
Ht TNR Ga Al Cr
Legible 4.1467 4.1067 4.0267 4.0000 3.9733
Ga Hr Cr Jm SITC
Attractive
3.1733 3.1333 3.1200 2.9067 2.9067
Jm SITC Hr HP Cl
Creative
3.6400 3.4267 3.3333 3.1200 3.0400
TNR Cr Ht Ga Al
Formal
4.1733 3.9733 3.8533 3.8133 3.7600
Cl HP Jm SITC Pb
Sloppy 3.1200 2.8000 2.7867 2.7333 2.3467
Jm SITC Hr Cl CB
Relaxed
3.2933 3.1067 3.0000 2.9067 2.8933
Jm SITC CB Ga Cr
Friendly 3.4933 3.3200 3.2400 3.2267 3.2267
Ht TNR Al Cr Ga
Confident
3.8267 3.7200 3.6667 3.6000 3.5867

Table 5 Five typefaces that were the most associated with each of the ten
personality traits and their means

3.2 Correlation Analysis

To investigate the relationship between the twenty four typefaces related to each of
the ten personality traits, we performed correlation analysis on the survey data.

Correlation analysis involves assessing the strength of the relationship between two

interval or ratio variables. Based on [12], the aim of the correlation coefficient is to

determine:

26
 • whether there is a real relationship between two interval/ratio variables;

• the direction of the relationship and

• the strength of the relationship.

We used Pearson's Correlation coefficient to measure the strength of the linear

relationship between each two typefaces. And our survey data satisfy the assumptions

of calculating the Pearson's correlation coefficient, which include:

• linear relationship between two variables

• continuous random variables

• both variables must be normally distributed

• two variables must be independent of each other

No Linear Relationship
-1.0 -m O m - 1,0
Perfect Negative Perfect Positive

Figure 12 Pearson correlation coefficient

If Pearson's correlation coefficient value is near ± 1, then it is a perfect

correlation. When Pearson's correlation coefficient value lies around zero, then there

is no correlation.

By performing correlation analysis on the survey data, we obtained a series of

correlation coefficients. In our study, a high positive correlation coefficient between

two typefaces indicated that participants perceived these two typefaces have very

similar personality traits; a high negative correlation coefficient between two

27
typefaces reflected that participants perceived these two typefaces have very

dissimilar personality traits.

A number of relatively strong correlations are found in our calculations. Table 6 is an

example and shows the Pearson's Correlation coefficients between each pair of

twenty four typefaces related to personality trait "Legible". Pearson's Correlation

coefficients between each pair of twenty four typefaces related to ten personality traits

are shown in Appendix E.

By examining the correlation coefficients in the matrix of all the personality traits, we

found some relatively significant correlations, which means there is a high degree of

correlation between these two typefaces. In the social sciences, a correlation of 0.30

using individual level data is considered a "good" correlation; a correlation above

0.40 is considered "strong" [13]. In our study, in order to reduce the number of

typefaces for further analysis, we set 0.60 or more as a "strong" correlation threshold.

We highlighted all the correlation coefficients that equal or greater than 0.60 in the

Table 6.

28
"3d
I-!

et

Cl

Q.

"S
U
(?
<U
(J

S3
S
a
_?
"3
—

1-
U
?
U

?
e2
We summarized the fifteen typefaces that exhibited strong correlations with the ten

personality traits. They are: Garamond, Belwe Lt BT, Harry Potter, Centaur,

Jokerman, Times New Roman, Arial, Chiller, Helvetica, Bauhaus 93, Kabel,

Rockwell, Snap ITC, Harrington and Footlight MT Light.

We compared results from the correlation and univariate analyses result from section

3.1. We found that most typefaces that appeared in the three typefaces that most

associated with the ten personality traits (Table 5) also exhibited strong correlations

with the ten personality traits. Only two typefaces, Cooper Black and Playbill, we

found them did not produce statistically significant results after examining their

correlation analysis results. Therefore these fifteen typefaces were used in further

analyses. The other nine typefaces, Cooper Black, Berlin Sans FB, Bernard MT

Condensed, Playbill, Poor Richard, Broadway, Kino MT, Impact and Onyx were not

included in the next factor analysis because they were not found to be the most

associated with the ten personality traits or they did not produce statistically

significant results in the correlation analysis.

3.3 Factor Analysis

We performed factor analysis by using Principal Components Analysis with Varimax

Rotation (Kaiser Normalization) to identify the common underlying factors between

typefaces and personality traits. Factor analysis is similar in reasoning to cluster

analysis. Generally, a factor analysis goes through two stages: deriving the factors,

30
then rotating them to enhance their interpretability.

3.3.1 Kaiser-Meyer-Olkin Measure of Sampling Adequacy

In order to detect whether or not a factor analysis was applicable with our survey data,

first, we performed Kaiser-Meyer-Olkin Measure of Sampling Adequacy on the rating

scores of the fifteen typefaces and ten personality traits to test if these typefaces could

be grouped into a smaller set of underlying factors.

We found that the Kaiser-Meyer-Olkin values of each personality trait for the fifteen

studied typefaces all produced values greater than 0.730. High Kaiser-Meyer-Olkin

values (close to 1.0) generally indicate that a factor analysis may be useful. If the

value is less than 0.50, the results of the factor analysis will not be very useful.

Therefore our data clearly supported the use of factor analysis.

3.3.2 Deriving Principal Component

Principal Components Analysis (PCA) was used to combine multiple correlated

variables into components. This method was used in the typeface persona studies

conducted by Bartram [14]. Bartram describes PCA as follows:

This is a technique for summarizing the inter-correlations between a large number

of scales in terms of a smaller number of independent factors. If four scales ... all

measure the same semantic property, principal components analysis should

31
 produce one main factor which accounts for most of the variation in the
correlations (i.e. the inter-relationships) between the four scales. This factor can

be taken to represent the semantic dimension which each of these scales is

measuring.

We used the correlation matrix of fifteen typefaces that we obtained from our

previous correlation analysis. The initial component structure started with as many

components as there were items in the analysis. In our case, we started with fifteen

items. Next, we calculated the eigenvalues, which represent the amount of variance in

the data, and explained the factor with which it was associated. The components were

extracted in order of the amount of variance that they explained. Therefore, the first

component had the highest eigenvalue, the second the next highest, etc. The first few

components explained the majority of the variance with the last few explaining only a

very small proportion of variance. We retained components whose eigenvalues

exceed 1 . Then, after the appropriate number of components have been determined, a

component matrix was calculated. This matrix identified the relationship between

variables and components.

As a general guideline, component loadings greater than 0.40 indicate that an item is

related or associated with a given component. We sorted the matrix and suppressed

values that were less than 0.40. We knew that if no relationship existed between the

variables then each variable would make its own unique component.

32
3.3.3 Varimax Rotation

After obtaining the components, we used Varimax rotation to enhance the

interpretability of each component and sort data until specific groups are indentified.

Varimax rotation is a change of coordinates that maximizes the sum of the variance of

the loading vectors. That is, it seeks such a basis that most economically represents

each individual - that each individual can be well described by a linear combination of

only a few basis functions. The rotation can aid in simplifying the factor interpretation.

Table 7 is an example showing the rotated component matrix of these fifteen

typefaces related to personality trait "Cheerful". Four components (factors) are

identified in this factor analysis calculation. Rotated component matrix of fifteen

typefaces related to the ten personality traits are shown in Appendix F.

33
 Rotated Component Matrix
Component

TNR Ch .905

FL Ch .859

Cr Ch .852

Ga Ch .756

Rw Ch .682

Ht Ch .654

Al Ch .622

SITC Ch .773,
Hr Ch .669!

Bh93 Ch .640 .600

Cl Ch .753

HP Ch .730

Jm Ch .545 .623

Kb Ch .826

BLB Ch .458 .508

Table 7 Rotated component matrix of fifteen typefaces related to the

personality trait "Cheerful"

3.3.4 Interpretation of Factors

Our factor analysis results revealed that three or four independent factors both

accounted for 50% of the total variance. We can thus categorize fifteen typefaces into

three or four groups. On closer examination of the factor analysis results, the ratings

of the fifteen typefaces and the values of their correlation, we finally decided on four

groups. Typefaces within a group correlated highly with the other typefaces in that
34
 group, and did not correlate highly with typefaces in other groups.
Group 1 - Garamond, Centaur, Times New Roman, Arial, Helvetica, Rockwell and
Footlight MT Light

Group 2 - Belwe Lt BT, Bauhaus 93 and Kabel

Group 3 - Jokerman, Snap ITC and Harrington

Group 4 - Harry Potter and Chiller

Items that had higher factor loadings were being more representative of the factor
than items with lower factor loadings. For example, all typefaces that correlated

positively in the group 3 rated much higher than those in the other three groups; thus,
cheerfulness was a common property of the typefaces in group 3 and was a
characteristic that distinguished those typefaces from the typefaces in the other groups.
Although typefaces in group 3 also rated highly on "Attractive", "Creative",

"Relaxed" and "Friendly", the differences between the groups were not as great, so
these characteristics were not considered to be the distinguishing characteristics. We
incorporated, compared and combined the rating scores of each personality trait
across the four groups and ranked them within each group. In summary, we labeled
the four groups based on these rankings as (Table 8).

35
Group Factor Typeface

Garamond, Centaur, Times New Roman, Arial, Helvetica,

1 Directness
Rockwell, Footlight MT Light

2 Gentleness Belwe Lt BT, Bauhaus 93, Kabel

3 Cheerfulness Jokerman, Snap ITC, Harrington

4 Fearfulness Harry Potter, Chiller

Table 8 Four groups and their corresponding typefaces

3.4 Multidimensional Scaling (MDS) Analysis

Multidimensional scaling analysis was performed to validate the results from the

factor analysis. MDS analyzes the dissimilarity of data in a way that displays the

structure of the distance between data like a geometrical picture.

We performed MDS on rating scores of fifteen typefaces related to ten personality

traits. Figure 13 is an example showing the MDS analysis of the fifteen typefaces

related to the personality trait "Friendly". We found that typefaces that are similar are

represented by points that are close together, and typefaces that are dissimilar are

represented by points that are far apart. The fifteen typefaces can be combined into

four groups in the Figure 13. Typefaces within a group were close together. There was

no overlap between groups. Therefore, the MDS analysis revealed the fifteen typeface

groupings results comparable to those from factor analysis and MDS analysis

validated the results from factor analysis. MDS displayed the typeface groups by
36
dimensional descriptions and factor analysis generated the typeface groups by factor

descriptions.

Derived Stimulus Configuration

Euclidean distance model

JnuFd SITC_RJ
O O

Rw_Fd
t+_Fd BUB-Fd FL_Fd Ga_Fd
c O
O ° Al F<1-O o&_Fd
[Link] Fd
Bh93 Fd
O

a_Fd
o
HP_Fd
O

Dimension 1

Figure 13 MDS analysis of fifteen typefaces related to the personality trait

"Friendly"

3.5 Demographic Differences

We performed a one way Analysis of Variance (ANOVA) on the survey data to detect

if there is a difference between demographic groups based on gender. A one way

ANOVA can be used to determine if two or more independent random samples come

from populations with different means. Gender serves as the independent variable in

our analysis. Figure 14 is an example of the histogram of typeface Garamond

concerning ten personality traits in terms of gender.

37
 4.50
4.00
3.50
3.00
C 2.50
ci
2.00
?
1.50
1.00
0.50
0.00 ?
Sp Rx Fd

Personality Trait
IMaIe D Female

Figure 14 Histogram of typeface "Garamond" concerning ten personality traits

in male and female groups

In Figure 14, the ? axis indicates the ten personality traits while the y axis indicates

the average values of rating scores of each personality trait in the two groups: male

and female. We found no statistically significant difference between the responses of

male and female participants for all fifteen typefaces and ten personality traits by

calculating and comparing the rating scores' means of each typeface based on male

and female participants. Moreover, we used the difference among means of each

typeface related to ten personality traits and assessed them with a one-way ANOVA.

The results were all insignificant (p> 0.05). Gender, thus does not have a significant

and sizeable effect on perceptions of the fifteen typefaces' personality traits.

For the remaining demographic data (age and education background), the numbers of

participants within these groups were not sufficient for a valid analysis. The majority

of participants stated to be over 20 years of age and only several participants claimed

having Technical School/Higher Vocational School and Junior College/Technical

College education.

38
 In short, our analysis revealed no statistically significant effects related to

demographic factors.

3.6 Discussion

The results of our statistical analysis clearly supported that there is a clear and strong
relationship between perceived typefaces and particular personality traits.

We removed nine typefaces after our statistical analysis. These typefaces were
statistically insignificant. As a result and the total number of studied typefaces was
thus reduced from twenty four to fifteen. Through series of statistical analyses on
perceived personality traits, we categorized these fifteen studied typefaces into four
groups. The four groups each contained typefaces that are related by typographic
characteristics. These results can provide typeface designers with some use
suggestions in terms of the typeface choices to suit different personality expressions.
We also detected that there is no significant difference for the participants' response
influenced by the gender factor.

Participants in our study consistently described specific personality traits to certain

typefaces, which was consistent with those results from previous research on typeface
and their personality traits (see [15] and [16]).

In [15], Shaikh, Chaparro and Fox (2006) investigated the relationship between
certain personalities and various fonts through an online survey. They concluded that
users consistently attributed personalities to fonts. The twenty fonts chosen for their

39
survey are Cambria, Constantia, Corbel, Candara, Calibri, Consolas, Times New

Roman, Arial, Verdana, Comic Sans, Century Gothic, Courier New, Impact, Gigi,

Kristen, Rockwell Extra Bold, Rage Italic, Agency FB, Georgia and Monotype

Corsiva. They used fifteen personality adjective pairs and semantic differential scales

in the font survey. They analyzed the top three fonts of each personality traits and

finally presented five groups (All purpose, traditional, happy creative, assertive bold

and plain).

Compared with their survey, we found that:

1. Three commonly used fonts, Times New Roman, Arial and Impact were also

chosen in our study.

2. We also used Creative, Attractive, Happy and Formal personalities in our survey

as compared to Shaikh's.

3. However, due to the difference in specific typefaces, personality traits and rating

scales used in our survey as compared to Shaikh's and former studies, we cannot

directly compare our study results with theirs and others.

Tables 9 and 10 display the study method and results compared between Shaikh's

(2006) and our studies.

Some issues that were not addressed in our study but need further investigation are

whether (a) the participants' age and educational background, (b) reading of familiar

or unfamiliar typefaces, (c) reading time, affect participants' responses.

40
 I ? ^
C/3
ß* .3
a o g ti
D.
t?
öS O
>
Pu

u
H

tí J2 S 3 a-
O
L.
V
S -3 Ja "è § t,
DO 3 «i c3 m la S
JJ U (D Ç Ii
P- U U Pu J < U (? »Í

e
S
M
o
s CJ
?:
o

u R U
o
P4 ¦s
H
"3 CA

S *H4 "¦» .58 O?

s CQ
-ft-

S:
CO

8 s LU -g. S
o
O
Pm
¦S
C S E^
V
Q. s
? to O
H
o g & sa ??
ci o
W ? e
C 0 m t.
(0 s u

I
?
C o 4>
2
*0 o ü O
TJ u
V-
C O
£ .S 0) T>
F O 03 CS
H < > GUuI
U 0
?
-o CO
.s ?» fa C
0
S Q ü
.2 S nj V
e 05
a; re
?3
?-«
O a
o U
Pi
0
—
ro >
-^ 3-
ft
E o
« d d O o O öS 4ft
u u u

O "3
a O 3
c/i C/5

3
O
 Group Typeface
Calibri, Century Gothic, Arial, Verdana, Corbel,
All Purpose
Candara

Traditional Constantia, Georgia, Times New Roman,

Cambria
Shaikh
Gigi, Kristen, Rage Italic, Monotype Corsiva,
(2006) Happy Creative
Comic Sans

Assertive Bold Impact, Rockwell Xbold, Agency FB

Plain Courier New, Consolas

Garamond, Centaur, Times New Roman, Arial,

Directness
Helvetica, Rockwell, Footlight MT Light
Our Gentleness Belwe Lt BT, Bauhaus 93, Kabel
Studies
Cheerfulness Jokerman, Snap ITC, Harrington
Fearfulness Harry Potter, Chiller

Table 10 Comparisons of study results between Shaikh (2006) and our studies

42
Chapter 4 Typeface Characteristics
Analysis

In this chapter, we describe typeface characteristics based on typographical and

aesthetics design of the fifteen typefaces, which have been divided into four groups

(Table 11), and provide some suggestions on typefaces' uses. Typographical design

characteristics are objective descriptions of typefaces, while aesthetic design

characteristics are subjective descriptions of typefaces, which mostly rely on viewers'

perception.

Group Typeface

Garamond Centaxir Times New Roman

Directness

Arial Helvetica Rockwell Footlight MT Light

Gentleness BelweLtBT liouhou/9ï Kabel
Cheerfulness ìFolerrnasn %ñup ITC "Harrington
Fearfulness
Hort/ p«ííef chill«
Table 11 Fifteen studied typefaces in four groups

4.1 Typographical Design Characteristics

First, we analyzed typographical design characteristics of the four groups that include

all fifteen typefaces. Typographical design characteristics include legibility, Serif and

Sans Serif, x-height, ascender and descender, etc. We will now discuss these
43
characteristics individually in detail.

4.1.1 Legibility

Legibility is one of the primarily concerns of typeface designers and is an important

part of typeface design. We examined the legibility of fifteen typefaces by analyzing

the survey data of personality trait "Legible". We calculated the means of rating

scores of personality trait "Legible" for all fifteen typefaces which have been

categorized into four groups, Directness, Gentleness, Cheerfulness and Fearfulness.

Group Typeface Legibility

Garamond 4.0137
Centaur 3.9726
Times New Roman 4.0822
Directness Arial 3.9726
Helvetica 4.1233
Rockwell 3.7123
Footlight MT Light 3.5342
Belwe Lt BT 3.3562
Gentleness
Bauhaus 93 2.4521
Kabel 3.2603
Jokerman 2.4658
Cheerfulness Snap ITC 2.3836
Harrington 2.7397
Harry Potter 2.0822
Fearfulness
Chiller 2.2466

Table 12 Mean values of rating scores of personality trait "Legible" for fifteen
typefaces within their corresponding groups

44
 No. Group Legibility
Directness 3.916
Confident 3.023
Cheerfulness 2.530
Fearfiilness 2.164

Table 13 Legibility comparison of four groups

From the mean values of personality trait "Legible" of fifteen typefaces and four

groups shown in Table 12 and 13, we have found that:

1) The most legible group is Directness, followed by the group Gentleness and

Cheerfulness; the most illegible group is Fearfulness.

2) Typefaces in the Directness group all have high values (>3.5) for the personality

trait "Legible".

3) The total value for the group Directness is much higher than the value of group

Fearfulness.

4.1.2 x-height Proportion

The typographical structure of text lines is determined from the vertical projection

profile, VP [17], as shown in Figure 15. Each component VP[i] represents the sum of

black pixels of the scardine i. The ul and bl scanlines, which estimate the upperline

and the baseline, correspond to the main peaks of VP, such that:

ul = i if? e[to,to+-\bo-to\]&max(VP[i + l]-VP[i]);

bl = i ifi G[to + -\bo-to\,bo]&max(VP[i -l]-VF\i]) .

45
? - height -bl-ul

wordheight = bo-to

? - height Pr oportion = ? - height / wordheight

x-height

Figure 15 Four typographical lines from vertical projection profiles

We examined the x-height proportion of fifteen typefaces included in the four groups

(Table 14). The word test samples were set in 36 points and converted to binary

images at 200*200 dpi resolution. The word test samples of fifteen typefaces are

displayed in Appendix G. The same test samples were used in x-height proportion,

ascender proportion, descender proportion and weight detection.

46
 Directness
No. Typeface x-height/word height
Garamond 0.4444
Centaur 0.3916
Times New Roman 0.5038
Arial 0.5571
Helvetica 0.5603
Rockwell 0.5145
Footlight MT Light 0.5036
Gentleness
No. Typeface x-height/word height
Belwe Lt BT 0.5639
Bauhaus 93 0.5180
10 Kabel 0.5644
Cheerfulness
No. Typeface x-height/word height
11 Jokerman 0.3736
12 Snap ITC 0.5347
13 Harrington 0.5180
Fearfulness
No. Typeface x-height/word height
14 Harry Potter 0.4138
15 Chiller 0.2917

Table 14 x-height proportions of fifteen typefaces in 36 point

x-height proportions of 1 5 typefaces (36 pt)

c
O 0.6
? 0.5
o
a 0.4
o
0.3
JC 0.2
'f 0.1
S. M
? 0
X

^ y S ¦*«¦
<?
<* # s° h* #v
^

Figure 16 x-height proportion of fifteen typefaces in 36 point

47
For the x-height proportions of fifteen typefaces shown in Table 14 and Figure 16, we
found that:

1. x-height ratios of the typefaces in group Directness are within the range of
0.40-0.56.

2. x-height ratios of the typefaces in group Gentleness are comparably larger (>0.50)

than all other groups. Typeface Kabel has the largest x-height ratio of 0.5644 as

compared to all other typefaces.

3. x-height ratios of typefaces in group Cheerfulness are within the range of

0.37-0.52.

4. x-height ratios of group Fearfulness are the smallest of the four groups. The

typeface Chiller has the smallest x-height ratio of 0.2917 in fifteen typefaces.

4.1.3 Ascender Proportion

Based on Figure 15, we find that:

Ascender -ul-to

Ascender Pr oportion — Ascender I wordheight

We examined the ascender proportion of all fifteen typefaces within their four groups

(Table 15).

48
 Directness
No. Typeface Ascender/height
Garamond 0.2741
Centaur 0.3147
Times New Roman 0.2556
Arial 0.2143
Helvetica 0.2057
Rockwell 0.2391
Footlight MT Light 0.2555
Gentleness
No. Typeface Ascender/height
Belwe Lt BT 0.2180
Bauhaus 93 0.2374
10 Kabel 0.2331
Cheerfulness
No. Typeface Ascender/height
11 Jokerman 0.3516
12 Snap ITC 0.2292
13 Harrington 0.2518
Fearfulness
No. Typeface Ascender/height
14 Harry Potter 0.3241
15 Chiller 0.3542

Table 15 Ascender proportions of fifteen typefaces in 36 point

sssísnateFí^giM pf©jî@rîisjîs sf ^S Sfptñcsss ?® p£|

*. 0.4
¦§) 0-35
•f 0.3
5 0.25
F 0.2
"s
F
0.15

Sj 0.05
re
O.i

0 111
.^ et? <y a*

¦<? </

Figure 17 Ascender/height of fifteen typefaces in 36 point

49
From Table 1 5 and Figure 1 7 we find that:

1. The ascender ratios of typefaces in group Directness are within the range of

0.20-0.32. Typeface Arial has the smallest ascender ratio (0.2143) of all fifteen

typefaces.

2. The ascender ratios of typefaces within the group Confident are very close to each

other and are within the range of 0.21-0.23.

3. The ascender ratio range of typefaces in group Cheerfulness and Fearfulness are

wide. Chiller, in group Fearfulness, has the largest ascender ratio of all fifteen

typefaces at 0.3542.

4.1.4 Descender Proportion

Based on Figure 15, we find that:

Descender - bo-bl

Descender Pr oportion = Descender I wordheight

We examined the descender proportion of fifteen typefaces within the four groups

(Table 16).

50
 Directness
No. Typeface Descender/height
Garamond 0.2815
Centaur 0.2937
Times New Roman 0.2406
Arial 0.2286
_5_ Helvetica 0.2340
Rockwell 0.2464
7 Footlight MT Light 0.2409
Gentleness
No. Typeface Descender/height
Belwe Lt BT 0.2180
Bauhaus 93 0.2446
10 Kabel 0.2025
Cheerfulness
No. Typeface Descender/height
11 Jokerman 0.2747
12 Snap ITC 0.2361
13 Harrington 0.2302
Fearfulness
No. Typeface Descender/height
14 Harry Potter 0.2621
15 Chiller 0.3542

Table 16 Descender proportions of fifteen typefaces in 36 point

descender/height proportions of 1 5 typefaces (36 pt)

as 0.35
<B 0.3
t 0.25

Ml
Q) 0.2
¦D
C 0.15
0> 0.1
?
W 0.05
¦8 o

ti»
/\/y ¿sAf ./y /y/// s
^
<&

Figure 18 Descender/height of fifteen typefaces in 36 point

51
From Table 16 and Figure 18, we find that:

1 . The descender ratios of typefaces in group Directness are close together and are

within the range of 0.22-0.30.

2. The desender ratios of typefaces in group Confident are in the range of 0.20-0.25.

Typeface Kabel has the smallest descender ratio of 0.2025 as compared to all

fifteen typefaces.

3. The descender ratios of all typefaces in group Cheerful are around 0.24.

4. The descender ratios of group Fearfulness are the largest one of all four groups.

Typeface Chiller in this group has the largest descender ratio (0.3542) of all

fifteen typefaces.

4.1.5 Weight Detection

The weight of font is reflected by the density of black surfaces on the white

background. This density (dn) is extracted from the horizontal profile P' ? .

Ph
Typex~~Typex Typexr Typex "3;
.LeJJm* lia yAia 1

Figure 19 Horizontal projection profiles

The weight is computed on the central part of the line located between

Hupper and Hbase , in order to be independent of the text line structure [17]. dn is thus

defined by:

52
 1 "
dn = — y]P'h[x]

We examined the font weights of fifteen typefaces within the four groups (Table 17),
in which the fifteen typefaces were set in 36 points in measurement.

Directness
No. Typeface Font weight
Garamond 26.2734
Centaur 19.9271
Times New Roman 34.0139
Arial 42.8067
Helvetica 32.8545
Rockwell 34.9534
Footlight MT Light 24.6890
Gentleness
No. Typeface Font weight
Belwe Lt BT 34.1732
Bauhaus 93 49.4027
10 Kabel 56.3408
Cheerfulness
No. Typeface Font weight
11 Jokerman 33.7589
12 Snap ITC 53.4029
13 Harrington 20.8273
Fearfulness
No. Typeface Font weight
15 Harry Potter 30.1218
16 Chiller 17.6524

Table 17 Font weight of fifteen typefaces in 36 point

53
 weight of 15 typefaces (36 pt)

S> 40

/vv v///y
<? <* &¿ *
&
¿sj-ss
s° ^ **V * ? ?
<f ^

Figure 20 Weight of fifteen typefaces in 36 point

From Table 17 and Figure 20, we find that:

1 . The weights of typefaces in group Directness are within the range of 20 to 43 .

2. The weights of typefaces in group Gentleness vary from 34 to 56, in which

typeface Kabel has the largest weight (56.34) of all fifteen typefaces.

3. The weights of typefaces in group Cheerfulness vary greatly from 20 to 53.

4. The weights of typefaces in group Fearfulness are comparably small.

Typeface Chiller has the smallest weight (17.65) of the fifteen typefaces.

4.1.6 Serif and Sans Serif

We classified all fifteen typefaces based on typographical features. The classification

is shown in Table 18.

54
 No. Group Typeface Classification
Garamond Serif
Centaur Serif
Times New Roman Serif
Directness Arial Sans Serif
Helvetica Sans Serif
Rockwell Slab Serif
Footlight MT Light Display Serif
Belwe Lt BT Display Serif
Gentleness Bauhaus 93 Display
10 Kabel Display Sans Serif
11 Jokerman Display
12 Cheerfulness Snap ITC Display
13 Harrington Display
14 Harry Potter Display
Fearfulness
15 Chiller Display

Table 18 Fifteen typefaces classified based on typographical features,

classification refers to [10]

From Table 1 8, we find that the typefaces in group Directness consist of Serif, Sans

Serif, Slab Serif and Display Serif. The typefaces in group Gentleness are Display
Serif, Display Sans Serif and Display. The typefaces in group Cheerfulness and

Fearfulness are all the Display typefaces. Display Sans Serif and Display Serif are

basic Sans Serif font and Serif font but some may be designed with only capital letters

or work best in larger sizes.

Round-shaped characters were selected as the conspicuous character patterns.

Conspicuous character patterns are special patterns designed for easier detection in

scenery images. We compared our fifteen studied typefaces with those conspicuous

character patterns defined in [18], and found some typefaces in the Gentleness and

Cheerfulness groups such as Typeface Bauhaus 93, Kabel and Snap ITC are

55
 round-shaped characters.

We use capital C to illustrate the differences between Serif, Sans Serif and Display
typefaces of our fifteen studied typefaces. The Capital Cs of the fifteen typefaces are
shown in Table 19. The order of typefaces in each group is the same as that in Table
1 1, similarly hereinafter.

Group Capital "C"

Directness
CCCCCCC
Gentleness
COC
Cheerfulness
CCC
Fearfulness
CC
Table 19 Capital "C" of fifteen typefaces in 36 point

[Link] Serif

For Serif capital C, there is a vertical or angled serif at the top and bottom of the bowl.
These two serifs are not normally identical and symmetrical. The lower end of C

could be a sharp point (see Times New Roman and Footlight MT Light in Table 19).
The upper end of serif capital C is identified by an implied spur (Centaur), bracketed
spur (Times New Roman) and slab serif (Rockwell), as shown in Figure 21.

56
 r> r^ r*
Figure 21 Upper end of Capital "C", implied spur, bracketed spur and Slab
Serif (from left to right: Centaur, Times New Roman and Rockwell)

[Link] Sans Serif

As displayed in Figure 22, for Sans Serif and Display Sans Serif typefaces, the end

stroke of capital C could be at any angle, oblique (Arial), horizontal (Helvetica) or

oblique and vertical (Kabel).

CCC
Figure 22 Ends of Capital C of Sans Serif (from left to right: Arial, Helvetica
and Kabel)

[Link] Display
Capital Cs of Display typefaces do not have regular round forms as do those of Serifs'

or Sans Serifs'. The arcs of Display typefaces are irregular and squeezed into the

vertical axis, (see C of Harry Potter in Table 19). In addition, the stroke weight of

capital C is not constant in most Display typefaces (Snap ITC, Jokerman and Chiller).

There are many litter stickers along the bowl of C in Jokerman. The cut of C in

Harrington even has flourished treatment (Figure 23).

57
 CCCcC
Figure 23 Capital Cs (from left to right: Harry Potter, Snap ITC, Jokerman,
Chiller and Harrington)

[Link] Legibility between Serif and Sans Serif

In the typographic literature [19], Serifs are generally believed to have a significant
impact on readability. There are two main reasons cited to explain why Serifs should
enhance legibility.

First, Serifs are believed to increase letter discrimination by making the spatial code

of letter forms more complex. Mclean [20] wrote: "Sans Serif type is intrinsically less

legible than seriffed type. . .because some of the letters are more like each other than
letters that have serifs, and so the certainty of decipherment is diminished."

Second, Serifs are thought to increase the visibility of the ends of strokes, increasing
the salience of the main strokes of the letters. Rubinstein [21] wrote: "Serifs have an

important role in the readability of type, providing. . .accentuation to the ends of

strokes that may help the reader read faster and avoid fatigue."

On the contrary, there are also some researchers who question if Serifs enhance

legibility. Moriarty and Scheiner [22] showed that there is no difference between Serif
and Sans Serif typefaces in terms of legibility. Also in [19], there are some reasons

supporting that Serifs may have little effect on legibility. Serifs are generally
ornamental rather than an essential part of the letter form. If they do affect legibility, it

58
might be reasonable to suppose that they interfere with letter recognition, since within
a simple letter-form template, they might simply act as a form of noise.
From our font survey results, the three most legible typefaces include Helvetica,
Times New Roman and Garamond with scores that are very close (Table 12), which

shows there is not a big difference between Serif and Sans Serif on legibility issues
within our present study. By contrast, the three most illegible typefaces include Harry

Potter, Chiller and Snap ITC (Table 12). The most legible group is group Directness

and the most illegible group is group Fearfulness. This may suggest that legibility is
diminished by the use of exaggerated ornamental elements and prominent

typographical features. Moderate typographical design characteristics increase

typeface legibility. The simpler a typeface design is, the more legible it is.

4.1.7 Character Stroke Contrast Design

Stroke contrast of characters is an important issue in typeface design. We examined

the ratio between left and right stem of capital "U" for all fifteen typefaces to

illustrate the stroke contrast design. The left stem of U is thick, and the right one is

thin, which is the tradition of Serif design. For Slab Serif and Sans Serif, the two

stems are always thick and symmetric. Some examples of capital U in typefaces used

in our studies are shown in Figure 24.

59
 UUUUUU
Figure 24 Examples of Capital U (from left to right: Centaur, Times New
Roman, Arial, Rockwell, Baubaus 93 and Harry Potter in 36pt)

Directness
No. Typeface Width ratio of left /right stem of U
Garamond 2.8
Centaur 3.5
Times New Roman 2.5
Arial 1.0625
Helvetica 1.0625
Rockwell 1.0625
Footlight MT Light 1.5
Gentleness
No. Typeface Width ratio of left /right stem of U
Belwe Lt BT 0.7667
Bauhaus 93 1.0303
10 Kabel 1
Cheerfulness
No. Typeface Width ratio of left /right stem of U
11 Jokerman 0.5807
12 Snap ITC 1.3333
13 Harrington 0.7059
Fearfulness
No. Typeface Width ratio of left /right stem of U
14 Harry Potter 1.1111
15 Chiller

Table 20 Ratios between the two stems of U for fifteen typefaces

From Table 20, we find that the ratios between the two vertical stems of U vary from

0.7 to 4 for the fifteen studied typefaces. Typeface Chiller has the strongest stroke

contrast of 4, which is the most exaggerated contrast of all fifteen typefaces. The

contrast of two stems in U for most Sans Serif typefaces is consistent and is not as

60
strong as Serif typefaces.

4.1.8 Character Width and Height Design

The ratio between width and height of characters is another important issue for

typeface design. We use the capital "O" to illustrate this ratio for our fifteen studied

typefaces. Capital Os of all fifteen typefaces are shown in Table 2 1 .

Group Capital "O'

Directness
OOOOOOO
Gentleness
ooo
Cheerfulness
0·0
Fearfulness
Oo
Table 21 Capital "O" in fifteen typefaces in 36 point

The form of the O impacts the design of all other round letters, such as Q, C, etc.

According to [3], the design of Serif O varied greatly over time. Early Serif O has

thick and thin strokes with oblique emphasis. Later Serif O becomes more oval and

upright, with both higher and lower contrasts. The Sans Serif O has fewer options

compared to its Serif counterpart, since the stroke contrast is not high, even the letters

are designed with equal stroke widths. This tradition is also present in other Sans Serif

61
character designs. For Display typefaces, the O is not in the traditional circular or oval
form, but is in the rounded rectangle, square, or asymmetrical form. All kinds of

random O forms, including rectangular, square, diamond, and triangular forms are

present in Display typefaces. The capital Os of Display typefaces do not keep the
traditional upright form. This characteristic also undermines the legibility and
readability of Display typefaces. The O of typeface Snap ITC is flat, as extends by
stretching and thus ruins the stroke weight and proportion. In contrast, capital O of

typeface Harry Potter is narrow, as it is condensed by squeezing the letter, and the
width of letter is not regular (Table 21). The capital Os' of Display typefaces are not

symmetric and are not round forms at all. Even the O of typeface Chiller is not a close
circle.

We calculated the ratios between width and height of capital O for the fifteen

typefaces. These values are shown in Table 22.

62
 Directness
No. Typeface Width/height of O
Garamond 1.0625
Centaur 1.0103
Times New Roman 0.9314
Arial 0.9189
Helvetica 0.9204
Rockwell 1
Footlight MT Light 0.9897
Gentleness
No. Typeface Width/height of O
Belwe Lt BT 1.019
Bauhaus 93 1.0097
10 Kabel 1.0317
Cheerfulness
No. Typeface Width/height of O
11 Jokerman 0.8475
12 Snap ITC 1.0727
13 Harrington 0.8019
Fea rfulness
No. Typeface Width/height of O
14 Harry Potter 0.5688
15 Chiller 0.7403

Table 22 Ratio between width and height of O for all fifteen typefaces

From Table 22, we find that:

1 . The capital O is not necessary a true circle in the typeface design, it is always in

an elliptical form. Capital O in Rockwell is the only true circle, as its ratio is 1 .

2. The ratios between width and height of capital O of typefaces in the Directness

and Gentleness groups are within the range of 0.9-1.1, close to a true circle.

3. The width and height ratios of typefaces in group Cheerfulness vary slightly, more

than those of typefaces in group Directness and Gentleness, as they are within the

range of 0.8-1.1. The capital O of typeface Snap ITC has the largest width and

63
 height ratio of 1.0727 of all fifteen typefaces.

4. The width and height ratios of typefaces in group Fearfulness are the smallest of

all four groups. The Os of typefaces in this group are therefore greatly narrower

than those in the other three groups. The O of typeface Harry Potter is the

narrowest with the width and height ratio of 0.5688.

4.1.9 Stem and Cap Height Design

Capital Y has two diagonal strokes and a vertical stem (Figure 25). The capital Ys of

the fifteen studied typefaces are shown in Table 23.

stem ^y caP
height Jl teigfr

Figure 25 Stem height and Cap height of Capital "Y"

64
 Group Capital "Y"

Directness
YYYYYYY
Gentleness
YYY
Cheerfulness
Yyy
Fearftilness
VY
Table 23 Capital "Y" for all fifteen typefaces in 36 point

Directness
No. Typeface Stem/Cap height of Y
Garamond 0.3830
Centaur 0.4316
Times New Roman 0.3878
Arial 0.4112
Helvetica 0.3925
Rockwell 0.4455
Footlight MT Light 0.3763
Gentleness
No. Typeface Stem/Cap height of Y
Belwe Lt BT 0.3333
Bauhaus 93 0.5455
10 Kabel 0.3609
Cheerfulness
No. Typeface Stem/Cap height of Y
11 Jokerman 0.2941
12 Snap ITC 0.4018
13 Harrington 0.2891
Fearfulness
No. Typeface Stem/Cap height of Y
15 Harry Potter 0.4183
16 Chiller 0.4862

Table 24 Ratios between stem and Cap height of Y of fifteen typefaces

65
Typeface designers are interested in the design of the stem height of Y. If the vertex is

too low, the Y will be top heavy. However, if the vertex is too high, the space between

the arms will be too small, and the gesture of the arms will look timid. According to

[3], in general, the stem of a Serif capital Y should fall between 35-50% of the capital

height.

We calculated the ratio between the stem and cap height of Y for the fifteen typefaces.

These ratios are presented in Table 24. From this table, we find that:

1 . The ratios between the stem and Cap height of typefaces in the group Directness

are within the range of 0.38-0.45.

2. The ratios between the stem and Cap height for typefaces in the Gentleness group

are within the range of 0.42-0.50, in which typeface Bauhaus 93 has the largest

stem/height ratio of 0.5455 of all fifteen typefaces.

3. The ratios between the stem and Cap height of the typefaces in group

Cheerfulness vary greatly. Typeface Harrington has the smallest stem height ratio
of 0.2891.

4. The ratios between the stem and Cap height for typefaces in the group Fearfulness

are within the range of 0.42-0.49.

66
4.1.10 Character Stroke Design
[Link] Stroke Length

Group Capital "E"

Directness
EEEE EEE
Gentleness
EGE
Cheerfulness
EEe
Fearfulness

Table 25 Capital "E" of fifteen typefaces in 36 point

We use capital E from all fifteen typefaces to demonstrate the stroke design

characteristic. Table 25 shows capital E for the fifteen studied typefaces. For capital E,

the lengths of the three arms in relation to the stem and to each other give the letter its

character [3]. Based on our results, we found that:

1. The capital E of Sans Serif typefaces, such as Arial and Helvetica in group

Directness and Gentleness, include three bars that are optically the same in length.

However, the lengths of the horizontal bars of Serif typefaces, such as Times New

Roman and Centaur in group Directness and Gentleness vary; the central bar is the

shortest, and the bottom is the longest.

2. The lengths of the three horizontal strokes of E of typeface Chiller in group

Fearfulness differ from each other and contrast greatly. Moreover, the lowest arm

67
 is much shorter than the upper one.

3. The central bar of E of Harrington in group Cheerfulness is the longest one of the
three bars, and all three bars are not horizontal and curvilinear.

4. The stress or bias of a roman font is the angle determined by the direction of the

thicker stem strokes [3]. The main stem of E in group Directness and Gentleness
is strictly perpendicular to the baseline.

5. By contrast, most typefaces in group Cheerfulness and Fearfulness are not

perpendicular to the baseline. The vertical stems of E of typeface Jokerman and

Chiller are slightly slanted to the right. The main stem of typeface Harrington is
curvilinear.

Similar features can be found in other letters in these fifteen typefaces that have main
upright strokes, such as B, D, K and R.

68
[Link] Stroke Form

Group Capital "D"

Directness
DDDD DDD
Gentleness
DDD
Cheerfulness
Dd?)
Fearfulness
DD
Table 26 Capital "D" of fifteen typefaces in 36 point

We use capital D as an example to illustrate the stroke form's design characteristic.

Table 26 shows the capital D for all fifteen studied typefaces. Based on our results, we

found that:

1 . The capital D of Sans Serif typefaces, such as Arial and Kabel in group Directness

and Gentleness, includes arcs that are almost semi-circular and are in a half round

shape.

2. The radian of the bowl of D of typeface Harry Potter in groups Fearfulness is

flatter than the radian of the bowl of D of typeface Arial in group Directness, and

it looks like a semi oval, squeezed in the horizontal axis.

3. The curvilinear stroke of D of typeface Snap ITC in group Cheerfulness is

irregular and squeezed in the horizontal axis. In addition, the curvilinear stroke

weight of D is not constant for typeface Snap ITC in group Cheerfulness and
69
 typeface Chiller in group Fearfulness.

Similar features of the fifteen typefaces can be found in other letters that have round

strokes, such as B, D, O and Q.

4. 1.10.3 Stroke Joining Part Treatment

Group Capital "M"

Directness
MMMMMMM
Gentleness
MmM
Cheerfulness
MMM
Fearfulness
M^
Table 27 Capital "M" of fifteen typefaces in 36 point

The treatment of the junction area where two strokes intersect in a letterform, such as

the vertices of M and N, is of interest. We use capital M to illustrate this design

characteristic. Table 27 shows the capital Ms for all fifteen studied typefaces.

1 . The vertex of the two diagonal strokes of M for most typefaces in the groups

Directness and Gentleness do not overshoot and are close to the baseline including

Serif and Sans Serif typefaces, except the typeface Kabel in group Gentleness. In

this case, the sharp point vertex of the two diagonal strokes of M is below the
baseline.

70
2. The vertex of the two diagonal strokes of M of typefaces in groups Cheerfulness

and Fearfulness protrude conspicuously above the baseline. The Vertexes of M of

typefaces in group Cheerfulness vary in form. The vertex of M of typeface

Harrington is a loop circle.

4.1.11 Counter Design

Group Lowercase "b"

Directness
bbbbbbb
Gentleness
bbb
Cheerfulness
bbb
Fearfulness
H
Table 28 Lowercase "b" of fifteen typefaces in 36 point

We chose lowercase "b" to illustrate the different counter designs of the fifteen

typefaces we studied. Table 28 shows the different designs of lowercase "b" for all

fifteen typefaces. We found the following results:

1 . In typefaces in groups Directness and Gentleness, counters of character b appear

even and in circular form.

2. In typeface Snap ITC of group Cheerfulness, lowercase b has a very small counter.

Moreover, due to the inconsistent weight of the round stroke, the closed counter is

71
 not in the center of the letter, but is in the upper part. In typeface Jokerman of

group Fearfulness, lowercase b has a spiral counter. The counter of lowercase b of

typefaces in group Fearfulness are not in a circle form and lowercase b of typeface
Chiller does not have a closed counter.

4.1.12 Character Space

In [4], Walter wrote: "the success or failure of a type is very much a question of

getting a good balance of white inside and outside the letters. The interior areas of

letters are fixed by the shape of the letters, but the spaces at both sides of them are at
will."

In order to examine whether the character space influenced participants' attitudes

toward typefaces, we chose some standard characters, including capitals "H", "O",

and lowercases "h" and "o" by using Fontlab 4 to detect the inter-letter space. Table

29 shows the left and right side bearings of these standard characters for all fifteen

studied typefaces. We found that the inter-letter settings of the characters differ

considerably from one typeface design to another. For typeface Kabel of group

Gentleness, the side bearings of the four standard characters are much bigger than

those of the other typefaces. On the contrary, for typeface Snap ITC of group

Cheerfulness and typeface Harry Potter of group Fearfulness, the left side bearings

have negative values and are smaller than others. From these values, we found

negative values appear only in the typefaces belonging to groups Cheerfulness and

72
 Fearfulness.

Directness
H
No. Typeface
(LSB1RSB) ([Link]) (LSB1RSB) (LSB1RSB)
Garamond (39,53) (94,98) (30,27) (73,74)
Centaur (31,55) (84,80) (29,11) (78,78)
_3_ Times New Roman (35,42) (72,79) (13,13) (69,71)
Arial (164,165) (99,92) (135,139) (68,76)
J_
_5_ Helvetica (158,156) (80,80) (133,133) (72,72)
Rockwell (55,53) (102,101) (42,12) (78,78)
_6_
1 Footlight MT Light (61,63) (76,75) (50,42) (42,45)
Gentleness
Belwe Lt BT (102,103) (72,72) (94,93) (68,67)
Bauhaus 93 (137,136) (55,54) (109,109) (48,47)
10 Kabel (209,184) (242,197) (217,242) (242,176)
Cheerfulness
11 Jokerman (141,114) (102,102) (76,129) (64,64)
12 Snap ITC (-10,-10) (47,47) (-8,35) (37,37)
13 Harrington (12,124) (90,90) (83,152) (74,82)
Fearfulness
14 Harry Potter (8,-57) (33,31) (12,12) (12,7)
15 Chiller (74,103) (72,71) (23,90) (63,63)

Table 29 Left side bearings (LSB) and right side bearings (RSB) of characters
"H", "O", "h" and "o" for fifteen typefaces

4.2 Aesthetic Design Characteristics

In addition to typographical design characteristics, we analyzed the aesthetic design

characteristics of all fifteen typefaces in their four groups. According to some

marketing research studies on logo design ([23] and [24]), three universal aesthetic

dimensions of graphic logo design were proposed, elaborateness, naturalness and

harmony. Since many of our studied typefaces are Display typefaces, we tried to use

73
 these three aesthetic design dimensions to examine our four studied groups and their

corresponding typefaces. These three dimensions were analyzed individually and will
be discussed in detail.

4.2.1 Elaborateness

Elaborate is not simply intricate, but appears to capture the concept of design richness
and the ability to use simple lines to capture the essence of something [23].

Elaborateness is composed of several design characteristics: ornament vs. briefness,

depth vs. flatness and special use vs. common use.

[Link] Ornament vs. Briefness

The design of typefaces in group Directness is brief, simple, and is without any
ornaments. The form of characters of typefaces in this group is standard and rigid. For
example, the capital Os of group Directness are always circular or oval. Rectangular,
square, diamond and some other random O forms do not appear in group Directness.
Compared with the briefness of typeface design within the group Directness, some

special ornaments were incorporated in Display typefaces within the groups

Gentleness, Cheerfulness and Fearfulness. For example, the design of typeface
Jokerman in group Cheerfulness, includes some little stickers, small circles or even

star figures added deliberately to the main strokes of letters. These ornaments are

distributed evenly on strokes (Figure 26).

74
 Aa Life
Figure 26 Ornaments of typeface "Jokerman"

Stems and other strokes in the letterform of typeface Snap ITC of group Cheerfulness
include slanted, convex, and concave lines (Figure 27). The special curly strokes
communicate a feeling of cheerfulness.

Ad Lit»
Figure 27 Curvaceous strokes of typeface "Harrington"

[Link] Depth vs. Flatness

Depth gives the appearance ofperspective or a three dimensional design [23].

Typefaces in group Directness are flat and seldom have structure variation; the strokes
of these typefaces are always vertical and horizontal.

Typefaces in group Gentleness have a slight structural variation. For example, the
serifs in typeface Belwe Lt BT in this group are slanted and are in a ribbon form
(Figure 28).

rvwy
Figure 28 Curvaceous strokes of typeface "Belwe Lt BT"

Typefaces in group Cheerfulness and Fearfulness represent a wide range of structure

75
variation. These variations include stroke form, character size and proportion, etc.,

which add depth to the typeface and make it more distinctive.

[Link] Special Use and Common Use

Typefaces in group Directness are commonly and widely used in small sizes as text

typefaces for newspaper, textbooks, magazines, etc. They can also be used in large

sizes for Display typefaces, such as headings in advertisements. However, typefaces

in group Gentleness, Cheerfulness and Fearfulness are only suitable for headings in

large sizes.

4.2.2 Naturalness

Naturalness is composed of several design characteristics: organic vs. geometric and

printed vs. handwritten appearance.

[Link] Organic vs. Geometric

Organic designs are those that are made up of natural shapes, such as irregular curves.

Alternatively, geometric designs tend to represent less natural and more

synthetic-looking objects [23].

Typefaces in groups Directness and Gentleness are more geometric while typefaces in

groups Cheerfulness and Fearfulness are more organic in appearance. For example,

76
typeface Bauhaus 93 in group Gentleness, displays a rigid letterform style, which is

geometric and even (Figure 29).

fid lib
Figure 29 Geometric and even strokes of typeface "Bauhaus 93"

For typeface Harrington of group Cheerfulness, tight curlicues are added to the serifs,

no matter if they are capital or lowercase letters (Figure 30). The tight loop on the

terminal is the most identifying characteristic of typeface Harrington and makes it

more casual and original. It creates visual interest and provides a fun and vivacious

feeling.

fldlsib
Figure 30 Curly Serifs of typeface "Harrington"

[Link] Printed vs. Handwritten Appearance

For typefaces of groups Directness and Gentleness, their strokes are straight lines, and

their structure is rigid. The characters rest on the same baseline.

However, typefaces in groups Cheerfulness and Fearfulness, include letterforms that

look more handwritten and random. For the letterform of typeface Chiller of group

Fearfulness, all strokes are in handwritten form and convey shivery chilled feelings

77
(Figure 31). The great effect creates visual interest and provides a fearful feeling.
Moreover, there is no horizontal and perpendicular line in these two letterforms.

Figure 31 Shivery chilled strokes of typeface "Chiller"

4.2.3 Harmony
Harmony is composed of several design characteristics: symmetry vs. asymmetry, and
balanced vs. unbalanced.

[Link] Symmetry vs. Asymmetry

Symmetric designs appear as reflections along one or more axes. That is, the elements

on one side of the axis are identical to the elements on the other side [23].

The letterform design of typefaces in groups Directness and Gentleness exhibit

symmetry everywhere, and the symmetrical design generally lends the letterform to a

more formal appearance. On the contrary, asymmetry is a more common letterform

design of typefaces in groups Cheerfulness and Fearfulness.

[Link] Balanced vs. Unbalanced

Balance is related to symmetry because symmetric designs are normally considered

balanced. The reverse is not true, however (i.e., an asymmetric design is not

78
necessarily imbalanced) [23].

Letterforms of typefaces in groups Directness and Gentleness are well proportioned

and balanced, as opposed to letterform designs of typefaces in groups Cheerfulness

and Fearfulness, which are examples of unbalanced designs. They have different

baselines, proportions and flexible ornaments.

4.3 Summary of Typeface Characteristics

By analyzing the typographical and aesthetic design characteristics of our four studied

groups and their corresponding typefaces, we obtained some conclusions about

typeface design and the personality traits they convey. These conclusions are

summarized mainly from analyses of the individual characters.

4.3.1 Typographical Characteristics

We examined the typographical characteristics of our studied typefaces from two

aspects, groups and their representative typefaces.

[Link] Typographical Characteristics of Four Groups

We analyzed some typographical characteristics of our fifteen typefaces including

x-height, ascender and descender ratios, weight, serif design, character stroke contrast,

character width and height, stem height and cap height, stroke form, stroke length,
79
stroke joining part treatment, character space, and counter design. Table 30 shows the

mean values of four main typographical characteristics of our four study groups.

x-height ratio ascender ratio descender ratio font weight

Directness 0.4965 0.2513 0.2522 30.7883

Gentleness 0.5488 0.2295 0.2217 46.6389

Cheerfulness 0.4754 0.2775 0.247 35.9964

Fearfulness 0.3528 0.3392 0.3082 23.8871

Table 30 Mean values of four main typographical characteristics of our four

study groups

From Table 30 and our analysis we found that:

1 . The values of typographical characteristics of typefaces in group Directness are

moderate compared with the other three groups and it is the most legible groups

in four groups. It proved the balance between a moderately large x-height,

ascender and descender ratios is very important for typeface legibility.

2. The typefaces in group Gentleness have largest ratios on x-height and font weight,

while smallest ratios on ascender and descender ratios in four groups. Legibility

of the Gentleness group ranks second out of the four groups. The typefaces in

groups Directness and Gentleness are easy to read.

3. The values of typographical characteristics of typefaces in group Cheerfulness

and Fearfulness vary in a wide range. Typefaces that have minimum or maximum

values in our fifteen typefaces fall mainly within these two groups. Some

typefaces in these two groups have very flexible and exaggerated values,

80
 especially on ascender and descender ratios. This exaggeration creates visual

interest, making typefaces prominent and provides readers with strong visual

feelings. The typefaces of these two groups consist only of Display typefaces, and

are easy to catch readers' eyes compared with Serif and Sans Serif typefaces from

the Directness and Gentleness groups. Legibility of these two groups is worse

than the Directness and Gentleness groups.

4. There are trade-offs between typeface legibility and strong visual feelings

conveyed by typefaces. Specifically, moderate design increases the typeface

legibility, but decreases prominent responses. For example, typeface Helvetica in

group Directness scored very highly on legibility, low on creative and cheerful

and average on relaxed in our font survey. Typeface Jokerman in the Cheerfulness

group scored first on cheerful and creative and low on legible. The typefaces in

group Gentleness are less prominent compared to the other three groups. They

produced average scores on several typeface traits, such as cheerful, friendly,

confident and relaxed.

[Link] Typographical Characteristics of Groups' Representative

Typefaces

On the basis of the survey results and analysis of groups' typographical characteristics,

we select four typefaces that represented the characteristics for each of the four

81
groups.

• Directness Group and Typeface Helvetica

This group includes common, highly legible typefaces. Closer examination of the

ratings for each typeface within the Directness group, we found all typefaces in this
group rated highest than those in other groups on personality traits "Legible",
"Formal" and "Confident". Based on the ranking comparisons, we found five
typefaces, Helvetica, Times New Roman, Garamond, Arial and Centaur are in the

ranking of first five of all these three personality traits (Table 31).

Personality Typeface
Ht TNR Ga Al Cr
Lg 4.1467 4.1067 4.0267 4.0000 3.9733
TNR Cr Ht Ga Al
Fm
4.1733 3.9733 3.8533 3.8133 3.7600
Ht TNR Al Cr Ga
Cn
3.8267 3.7200 3.6667 3.6000 3.5867

Table 31 Five typefaces that were the most associated with personality traits
"Legible", "Formal" and "Confident"

82
Table 32 shows the mean values of four main typographical characteristics of these

five typefaces of the Directness group.

x-height ascender descender font

Typeface ratio ratio ratio weight
Ht 0.5603 0.2057 0.2340 32.8545
Directness TNR 0.5038 0.2556 0.2406 34.0139
Ga 0.4444 0.2741 0.2815 26.2734
Al 0.5571 0.2143 0.2286 42.8067
Cr 0.3916 0.3147 0.2937 19.9271

Table 32 Means of four main typographical characteristics of five typefaces in

the Directness group

Among the five typefaces, the typeface Helvetica has the largest x-height ratio,

smallest ascender ratio and scores first on personality traits "Legible" and "Formal".

The descender ratio of typeface Helvetica is smaller compared with other three

typefaces and only larger than typeface Arial. Therefore typeface Helvetica is a good

example to represent the Directness group. As we noted, typeface Centaur has the

smallest x-height ratio and font weight in five typefaces, and it scores fifth on

personality trait "legible". The result also proved that a moderately large x-height and

font weight are very important for a legible typeface.

Gentleness Group and Typeface Belwe Lt BT

This group is unlikely to be a very prominent one as compared to other three groups

because the typefaces in this group rated moderate on all the personality traits. After

careful examination we found the typefaces of this group ranked comparably higher
83
on "Cheerful", "Legible", "Creative", "Relaxed" and "Friendly". Typeface Belwe Lt

BT scored highest than other two typefaces on these personality traits. Therefore, we
choose Belwe Lt BT to represent this group.

x-height ascender descender font

Typeface ratio ratio ratio weight
Gentleness BLB 0.5639 0.2180 0.2180 34.1732
Bh93 0.5180 0.2374 0.2446 49.4027
Kb 0.5640 0.2331 0.2025 56.3408

Table 33 Means of four main typographical characteristics of typefaces in the

Gentleness group

Compared the typographical characteristics of these three typefaces (Table 33) with
typefaces in the Directness group, we found they all have larger x-height, smaller
ascender, descender and thicker weight as compared to typefaces in other groups.

• Cheerfulness Group and Typeface Jokerman

Personality Typeface
Jm SITC Hr BLB Ga
Ch
3.4533 3.2933 3.0133 2.6533 2.6000
Ga Hr Cr Jm SITC
At
3.1733 3.1333 3.1200 2.9067 2.9067
Jm SITC Hr HP Cl
Cr
3.6400 3.4267 3.3333 3.1200 3.0400
Jm SITC Hr Cl CB
Rx
3.2933 3.1067 3.0000 2.9067 2.8933

Table 34 Five typefaces that were the most associated with personality traits
"Cheerful", "Attractive", "Creative" and "Relaxed"

We found all typefaces that correlated in the Cheerfulness group rated highest than

84
 those in other groups on personality traits "Cheerful", "Creative" and "Relaxed"

(Table 34). In addition, the three typefaces scored high on personality trait
"Attractive". Typeface Jokerman scored first among the three typefaces in these four
personality traits. Therefore typeface Jokerman is a representative example of the
Cheerfulness group.

x-height ascender descender font

Typeface ratio ratio ratio weight
Cheerfulness Jm 0.3736 0.3516 0.2747 33.7589
SITC 0.5347 0.2292 0.2361 53.4029
Hr 0.5180 0.2518 0.2302 20.8273

Table 35 Means of four main typographical characteristics of typefaces in the

Cheerfulness group

Careful examination of the four main typographical characteristics of these three

typefaces (Table 35), we found their typographical ratios vary greatly. Typeface
Jokerman has the smallest x-height ratio and largest ascender and descender ratios
among three typefaces. It appears that special ornaments, unbalanced and

asymmetrical design are the reasons that typeface Jokerman was chosen as the most

creative and cheerful typeface.

85
• Fearfulness Group and Typeface Harry Potter

Personality Typeface
Hp Pb Cl BMC KM
Ff
2.9067 2.4800 2.3867 2.1067 2.0933
Jm SITC Hr HP Cl
Cr
3.6400 3.4267 3.3333 3.1200 3.0400
Cl HP Jm SITC Pb
Sp 3.1200 2.8000 2.7867 2.7333 2.3467

Table 36 Five typefaces that were the most associated with personality traits
"Fearful", "Creative" and "Sloppy"

We found that the typefaces in the Fearfulness group are rated as most fearful and

sloppy (Table 36). In addition, the two typefaces in the Fearfulness group scored high
on personality trait "Creative". Typeface Harry Potter scored first in the personality

traits "Fearful". Therefore typeface Harry Potter is the representative example of the

Fearfulness group.

x-height ascender descender font

Typeface ratio ratio ratio
Fearfulness
weight
HP 0.4138 0.3241 0.2621 30.1218
Cl 0.2917 0.3542 0.3542 17.6524

Table 37 Means of four main typographical characteristics of typefaces in the

Fearfulness group

As we noted previously, the four main typographical characteristics of these two

typefaces vary greatly (Table 37). Typeface Harry Potter has a comparably larger

x-height and smaller descender ratio, while typeface Chiller has the smallest x-height

86
in our fifteen typefaces. These exaggerated values make them illegible, but might be
effective in grabbing attention. The special ornaments used in these two typefaces
make the typefaces overly unpleasant. That is the reason typeface Chiller was rated as

the sloppiest one.

4.3.2 Aesthetic Characteristics

We evaluated aesthetic characteristics of the four groups based on three aspects:

elaborateness, naturalness and harmony. In our analysis we found that:

1. The typefaces of groups Directness and Gentleness are less complex and more
plain compared with those in the Cheerfulness and Fearfulness groups. Different
ornaments are used in all the typefaces of group Cheerfulness and Fearfulness,

while there are no ornaments in the typefaces of the Directness group, and some

subtle ornaments in typefaces of the Gentleness group. The applications of

typefaces in the Directness group are extensive while those in group Gentleness,
Cheerfulness and Fearfulness are only suitable for headings in large sizes.

2. The letterform of typefaces in the Directness and Gentleness groups are more

geometric and carefully set. Those of the Cheerfulness and Fearfulness groups are
more flexible and natural. There are almost no horizontal and perpendicular

strokes in typefaces of the Cheerfulness and Fearfulness groups.

3. The typefaces in the Cheerfulness and Fearfulness groups are asymmetrical and
unbalanced, while those in the Directness and Gentleness groups are symmetrical

87
 and balanced.

4. The most attractive groups are Directness and Cheerfulness, their typefaces

ranked highest on personality trait "Attractive".

5. There are some trade offs in the aesthetic aspects of typeface design. Symmetrical

and balanced designs increase friendly responses and typeface legibility but

decrease attraction and prominent responses. The use of ornaments always

influences and decreases typeface legibility.

Table 38 shows the different levels of aesthetic characteristics of our four study

groups. The Directness group is highly harmony but not elaborate and natural. The

typefaces in this group are all common and highly readable typefaces. The

Cheerfulness and Fearfulness groups are high on elaborateness and naturalness but

low on harmony. Such typefaces are mostly special used in the content of

advertisements. While the Gentleness group is average on elaborateness and harmony

compared with other three groups but low on naturalness.

Elaborateness Naturalness Harmony

Ornament Depth Organic Handwritten Symmetry Balanced

Directness Low Low Low Low High High

Gentleness Average Average Low Low Average Average
Cheerfulness High High High High Low Low

Fearfulness High High High High Low Low

Table 38 Different levels of aesthetic characteristics of four study groups

 Compared with previous research on font and personality traits ([15] and [25]), we not
only performed analyses of font survey results and obtained the font groups, also did

further research on the typographical and aesthetic characteristics based on our study
groups and their representative typefaces and suggested the potential association with
typeface design and their personality traits.

4.3.3 Appropriate Uses

Since specific typefaces are associated with particular personality traits, we need to

consider the responses that typefaces might create. Thus typefaces should be carefully
selected to ensure appropriateness for the meanings and occasions. For example,
Times New Roman is better than Harry Potter when we choose a typeface between
them for a business document. With regard to the survey results, we derive the
following conclusions on typeface uses:

• Directness Group

Typefaces in this group are legible, formal and confident, but unimaginative,
unemotional and unrelaxed. Therefore such typefaces are commonly used, all purpose
and especially appropriate for the content of official documents, reports and forms.

89
• Gentleness Group

Typefaces in this group are less prominent and scored average on all the personality

traits. In addition, the typefaces of the Gentleness group are more legible than

typefaces in the Cheerfulness and Fearfulness groups. However, with regard to the

noticeable ornaments used in these typefaces, they are more appropriate used in the

commercial advertising and headings than for textual contents. However, the feelings

they evoked might not be as intense as typefaces in the Cheerfulness and Fearfulness

groups.

• Cheerfulness Group

Typefaces in this group are rated as cheerful, attractive, creative and relaxed. Such

typefaces are generally best for evoking a pleasant tone in the commercial

advertisement and children's reading books.

• Fearfulness Group

Similar as typefaces in the Cheerfulness group, the typefaces in the Fearfulness group

are also best for evoking intense emotional feelings. The difference is typefaces in this

group are displeasing and cold. Such typefaces are generally used in the commercial

advertising for special effects. In addition, typefaces in the Cheerfulness and

90
Fearftilness groups are not very legible, therefore they are often printed in large size

and more appropriate for the headings than texts.

91
 Chapter 5 Conclusion

In this chapter, we summarize the contributions of this thesis and present some future

work in this research area. By conducting a designed font survey, the personalities of
twenty four studied typefaces were identified. The twenty four typefaces were

reduced to fifteen typefaces and four groups were clustered and defined through a
series of statistical analyses. Typeface characteristics, including typographical and
aesthetic aspects, for all fifteen typefaces were examined and analyzed in detail.

5.1 Summary

Since most studies on typeface design concentrate on typeface legibility and

readability, the main purpose of this thesis is to provide a new point of view. We
aimed to investigate the relationship between certain typefaces and particular

personality traits. Then, based on the identified personality traits the typefaces
conveyed, we analyzed their design characteristics.

Firstly, an investigation to address whether or not participants think that particular

fonts are associated with tangible personality traits and to what degree fonts can
convey personality traits was conducted. In our study, a designed font survey was
used. The personality traits that twenty four typefaces conveyed were identified after
comparison and examination.

Secondly, we conducted a series of statistical analyses on the survey data. The

92
 Statistical methods used include correlation matrix analysis, factor analysis,

multidimensional scaling and one way analysis of variance. We reduced the number

of studied typeface from twenty four to fifteen and categorized them into four

different groups, typefaces within a group correlated highly with the other typefaces
in this group. We labeled these four groups as Directness, Gentleness, Cheerfulness

and Fearfulness for further typeface characteristic analysis.

We investigated and evaluated the typeface characteristics of all fifteen typefaces in

the four groups for typographical and aesthetic design characteristics. Design features
of each of the four groups were analyzed. Moreover, based on characteristic

comparison of the four groups and their corresponding typefaces, conclusions and

suggestions were made based on the design of typefaces and the potential association
with personality traits.

5.2 Future Work

Based on research in the fields of psychology and typography design, our study offers
a systematic method of typeface design analysis in terms of the particular personality

traits the typefaces conveyed. Current work is an initial step, however, more research
is required. Future work may take into account the following considerations:
1 . In the future, the selection of personality traits that are used in research should be

pilot tested and examined in more detail to help make studied personality traits
more accurate and specific.

93
2. Due to the limitations based on our study's methodology, we must also address

some issues which may have influenced the participants' responses, including
factors such as participants' reading comprehension, reading time, familiarity with
studied typefaces. All these factors need further investigation. The distribution of
participants based on age and education background should also be taken into
consideration in the future.

3. The analysis of typeface design characteristics should be more profound and

characteristics used for analysis may be comprehensive, diverse and in greater
detail.

94
 References

1 . Soo C. Hostetier, "Integrating typography and motion in visual communication,"

Tech. report, Department of Art, University of Northern Iowa, Cedar Falls, 2006.
2. [Link]
3. Karen Cheng, "Designing typeface," Yale University Press, New Haven,
Connecticut, 2005.
4. Walter Tracy, "Letters of Credit," David R, Godine, Publisher, London, 1986,
pp. 30-32.
5. Albert T. Poffenberger, and [Link], "A study of the appropriateness of type
faces," Journal of Applied Psychology Vol. 7, 1923, pp. 312-329.
6. Hebert Spencer, "The visible word," 2nd edition, Lund Humphries in Association
with the Royal College of Art, New York, 1969.
7. Charles Kostelnick, David D. Roberts and Sam Dragga, "Designing visual
language: Strategies for professional communicators," Longman, New York,
1997.

8. Ronnie Shunshan and Don Wright, "Desktop publishing by design," Microsoft

Press, Belleuve, Washington, 1994.
9. Jadette Laliberte, "La typographie moderne: Consequence de la revolution
industrielle?," Communication et Langages Issue 72, No. 2, 1987, pp. 60-76.
10. "Guide to Microsoft fonts," DesignTutor, No. 8, UK. pp. 1-7.
11. [Link]

12. Robert L. Miller, Ciaran Acton, Deirdre A. Fullerton and John Maltby, "SPSS for
social scientists," Palgrave Macmillan, New York, 2002, pp. 160.
13. Stephen A. Sweet, Karen Grace Martin, "Data analysis with SPSS," third edition,
Pearson Education, Upper Saddle River, New Jersey, 2008, pp. 106-107.
14. David Bartram, "Perception of semantic quality in type: Difference between
designers and non designers," Information Design Journal Vol. 3, Issue 1, 1982.
pp. 38-50.

95
 15. A. Dawn Shaikh, Barbara S. Chaparro and Doug Fox, "Perception of fonts:
Perceived personality traits and uses," Usability News, February 2006, Vol. 8,
Issue 1, pp. 1-6.
16. A. Dawn Shaikh, Doug Fox and Barbara S. Chaparro, "The effect of typeface on
the perception of email," Usability News, January 2007, Vol. 9, Issue 1, pp. 1-7.
17. A. Zramdini and R. Ingold, "Optical font recognition using typographical
features," IEEE Transaction on Pattern Analysis and Machine Intelligence, vol.
20 No. 8, August 1998, pp. 877-882.
18. S. Uchida, R. Hattori, M. Iwamura, S. Omachi and K. Kise, "Conspicuous
character patterns," Proclo"1 International Conference on Document Analysis
and Recognition, Barcelona, Spain, 2009, pp. 1 6-20.
19. A. Arditi, and J. Cho, "Serifs and font legibility," Vision Research, Vol. 45, No.
23, 2005, pp. 2926-2927.
20. R. Mclean, "The thames and Hudson manual of typography," Vol. 1, UK:
Thames and Hudson Ltd, London, 1980.
21. R. Rubinstein, "Digital typography: An introduction to type and composition for
computer system design," Addison Wesley, Boston, Massachusetts, 1988.
22. S. E. Moriarty and E. Scheiner, "A study of close-set text type," Jounal of
Applied Psychology, Vol. 69, No. 4, 1984, pp. 700-702.
23. P. W. Henderson, and Joseph A. Cote, "Guidelines for selecting or modifying
logos," Journal of Marketing, Vol. 62, No. 2, 1998, pp. 14-30.
24. Pamela W. Henderson, Joan L. Giese and Joseph A. Cote, "Impression
management using typeface design," Vol. 68, No. 4, 2004, pp. 60-72.
25. Eva. R. Brumberger, "The Rhetoric of typography: The persona of typeface and
text," Technical Communication, Vol. 50, No. 2, 2003, pp. 206-223.

96
Appendix A Order of Typefaces in the
Survey

Sample 1 : Cooper Black

Sample 2: Berlin Sans FB
Sample 3: Bernard MT Condensed
Sample 4: Garamond
Sample 5: Belwe Lt BT
Sample 6: Playbill
Sample 7: Harry Potter
Sample 8: Centaur
Sample 9: Poor Richard
Sample 10: Jokerman
Sample 1 1 : Times New Roman
Sample 12: Arial
Sample 13: Broadway
Sample 14: Kino MT
Sample 15: Impact
Sample 16: Chiller
Sample 17: Helvetica
Sample 18: Bauhaus 93
Sample 19: Kabel
Sample 20: Onyx
Sample 21: Rockwell
Sample 22: Snap ITC
Sample 23 : Harrington
Sample 24: Footlight MT Light

97
Appendix B Sample of the Font Survey

98
 -o (U Si
¦a ö JU
ö ·5 £ tí
cd 3 bß
cd
?
-S
N
co
a
«s â -(U (U
bß bß
Ö .§ ^ <u .S
CO JS. cd tí
a? Vl S-H
* i-? cd
(U
(U ¦h=î 'cd
(L)
tí
>
CU
Oh
O
U(U ö
O :H
J=!
?^1
!-p
00
co
ö a, a ?
^
.ti
'S
(U a, co
p-» cd
Cd T3
cd ?3 CO 1) (U e
tí
(U
td
cd "C 9
O
(?
S-H
1 1 t..
ce tí
(U O J=!
O
S-I
bß
ö
Oh £
? o
cu
a,
§* -«
>-?
cd
-tí
O
co
c/3
cd
(U
cd
Oh
>-!
cd
Oh
.,_,
? W) -tí (U ed J=! Ö
O
(? O (U £ ü g u
tí S-i JS /<L> *rt
V)
ci
?
eO '-S% H-*
O
cd
co
?
co
>
i> ítí co
CO c*s 13 O U-i bß >
tí cd O Tí ? CU
? (U JU H-*
Sj=¡
O 13 U a co >UMJ
"* 7Á
a S JS
O-,
CO
S T3 (U
CO
C(J
tî CO
1-1
H-*
(U (U
H—»
CN bß
cd
r .S ^
O bß tí
CX O Ö (U
? cd co
Ci-I S-. CU
O S-I s-,
o is o
c<3 cd cu
o c/a Oh ? O Oh tí
— (U- (U (U J=! CU «
H-* H-*
Ö b S-! H-*
O O co
(N -a O

V) e O
bß (U g ¦> tí
cd § S -S
(U
S-.
cd o
Id
• i—4

tí
C
O
O
S-.
Oh ?
tí
^O? -Ch
?-·
.>-Í3+3 [Link]
o in
(U s-,
(U (U (U cd
?-? tí
(U be J3
-tí (U tí .SP (U S^ "o tí
O 'co
•5 =S
•^*
tí
'-?
£ rt o
JU 3 2
^ ?
(U (U
S -o co O
co ed Oh (U
tí" CU J^ co
^ 2
*^ H-*
O
co
(U
(U
C/3
H-*
tí
(U
Oh
??
CN J=!
tí (N ~
-?-»
·> ·°
cd

ed -tí cd . . cd C4-, CU > cu

(U (U Oh CU Oh O
(U S-.
•8 >?
(U fa tu"
•~* !>
tí
CU
N
Oh S CO
cd
(U
>
tí (U
bO
ed
¦e
ed
O
O cd e «s h
O cu o
>-H O S-H co Cu tí CV1 CU O co O
O to
O J=!
"? 1) P-,
(U H-*
O
cd
O co
W ^H
?co Ja
-S «
cu
bß S3 tí
ed
C co
co
O
CX)
CU
-tí tí
Cu cd
•e H-*
G^ s-, ?
Ö (U ed
>
?
o
s_
CU
J=!
H-* e " -? -ti cu
H-*
cd
in "?
V) bß
(U
(U
G?
"o
co
(U
O
?-,
(U
o
<hS
CU
J=!
•B
CU
CO
O
Dh
CO
§ 'S
= M
<U~
CU bß co
tí ? >
13 'd !_ cd O cd ti
>- cu bß S-I
T3 O Oh ^ Ö ö (U cu CU ? cu
(U J^H
O H-*
cu *-> <% ocd
H-*
cd e N —« O
e
Xl U-I co co cd cd rs (U
CU O
S-H co ° ·— h_>
tí
H-* -tí
(U O CU ? cu
co .-2 ·— <u ,O ? J=!
co O
H-*
-s
CU
tí
cd tí
e,
S-I cd
ed 'C H-*
cd ed X
C tí cu JS s ? cu (U
O >> .SP 53 co
H-* tí cu ? >
CO O
cd
a C ö Oh cd ^ 'S
X _G
O O
cd
tí cu cd (U '~" S co
ti
O Oh cd tí
O
JS (U
co CU ? -Oh Oh ed
JS (U "S co
H S-H
H-*
H-* cd .2 Td CV CU bß
(U cd S-I (U
O -tí „- s_
> tí
S-I ?-» cd CU tí >, cu -S^ co JS
cd Oh JO (U
i?
cu _tí
J=! ..-*-! O
(U (U co ed X
(U co O
co
(U >
ed O ed
O
cu
J=I S cd tí
co
(U
O
Oh
ju"
JS ¡u <H-H O H-J co tí co ? "Hh
CU O (U O e
>> '> Oh >
CO C/3
CU
ed (H-I CU ed j=l S-H
CU JS "o O
S-H ?-,
cd
X JS
O H^ J=! Oh H C
JD O (U H
 F
E
2 X X X
UJ

«
e
?
Vi
U
M cu
a.
> oc
S
JS
?» X
'%O ?
?
=s
a
¡a
xj
> >*
^1 0)
cu
>¦ G
ß
O
Q)
•?
X
?
s) > a
«? <? .?
c¿ => > Q) >
??? —
(U -C
</> jy ^
CL D. T- a»

a. ?- E ?£
UOF 3
O
? er >< 4) ? W

^ Q- v2 *? .
^" ??
ß
?
3 -?^

— O 5 a.
O
>. a»

? cu
¦ O JS

e ?
X ? X X X
CU
CT» °"J5 W Vi
Ol
?-
CX) J¿ (U (U
I— 4-J 4-1
a.
eu

?\
JZ
O eu
Xi
CT) W
__ LD X!

(L) ^ ?

"D CO C
CC)
(J G? eu
43
CQ JQ H
<C re ? O
O
JS
CJ
CU
Vi

cu

O) e
f > ?3 ?
?: 3 .5 OJ
f t U Q. e
a? RS e
?-
Q. JS ?
SZ ? ? O ¡? ?
O O CQ CC O
 f
E
S
W

e«
N a
o
an
U
Oi
Q.
eu
e s>

o
=3
N .2

D cu

>· E
ß ?
Í4 fJ O
?
•s
O
S
a

ûS 9 0 « Jg .?

Of ? **£
pu »
ti «eg 3 ?:
Dì
? Ob"
O
!>> <?
er ¦*-»

Z a S « 3
ß

S o Se« ?
a»
a>
J3
a
5to
d «ir S

wjs S a
« f «
£ 08 Ä ?
A
O Ä r- ?
«
fa M f
M cm «?
Q * a

N
O
H O

< «ß O

> 0) C
G3 CD
t ? > Q)
F ? ?- ?,
CD CO e a. IS
C
Cu ?
?: CU ? _0 ? O
O Ö CO O
 2>
i
JS
'S
ai

ß
O
(?
hl

Q.
M OJ)
a
X
> '??
X =s
,?

>
a O)
X ? •s
U ?
5
CA
B a
.?
F f
:> > 9 ?>
ai
Q ^ a a ? a»

CL * e £ P .2«
È .3
O .2*
?
O s
Z a * ö 5 S3
IU

3 a §. +J
H
2 ° 2 ? ?
-° ? ~
<
:* E ¦= >> >
3 N ^ a O
(?

£ -e
?? f J=
f (U

G" *-» *J D.
— OO
O
LL
UJ m
F
Û "?
U a
m Ol

<
O
O

¡?
«

F C
13 >
t 3 U
-9 a.
e Q. C
C
(D F
jc ? ? _? O
O LL. CJ U. 00 ?. O
<?
 f
E
8}?
OJ

«
a
o

eu
a.
ÖD
s
X

a»
JS

cu

> S
S f
O TJ
w O
2
S
,O
VS «12 'S
S « ?*¦
5?

ses
Cu ??. ö is 3
O
?:

K « S ^> <?
a
a»
S —· o
*^ Cm 5*«s

S ?ß S
5
^1S » '5 S =s s "?
tu
2£ as a>
S3 «-- l>* ce
<u
Vi
a.
^ -oc ^ a>
u
h M IA
<u
Xi
ß ? CO ¦*¦*
«
X!
^ « CM ¦*«*

k.
ce ß ^ a»

«a: ce ß e

O
O
X!

C«

3 (IJ XJ e
> cu
t 3 CD •s X3
F t Q. X
QJ CD e Q. ja e
QJ ?-
O co
?: ? ? cu O
O UL O CO cu O
rr>
 i
g
¦?
LU

a
?
<?

a a*
oc JZ
a D)

%? X

?.

?—,
f
> E
a Q)
N T)
H o
O
2
C/) a

Z. ?*
? ? C
<? *-; —*
a, A ti cu
<**

·— ? a s DO
O co
?? ? ? O
CO
ß
??

g Ol
a>
OÍ) £
? ? <
?)
^-Î fi a O
a>
rt il Hi »? Z
cu
Wl
a
?-? tu

O ?
??
*j
C3

Ul U
eu

Q (U
»?
a
? «
a>

CQ
O
O

CU

0) C
> f
1C J) > f
O
-9 E CL
a.
?
s e
f e
JZ O ?
O UL O CC O
 f
E
g
'S
UJ

e
o

a.
OX)
s CO
X
X
«2
JS

a>
X > E
S ?
O ?>
U ?

> o a
;> .o
o
Qí 1/1 O
-t—»
a </>
g· a 0 ¦*
?
a V-< *^

.3
JC

O CT «O ?O S° O
>>

Ou C ? 3 s

O
C JS
—3
<
CD
-t rv U s O
5 Kl 5 Ol
?
O)
U
Si o o a.
t-pt ··""»
P-"* 4—1 4—0 U

??

JQ
? ¦*¦»

m JS

a>

(?
S
«
GM a»
PQ J=

?
?

ce
di

F
Zi > F 3» F
T. 3 5? > ?? f ?
? t ? Q. ? ¦g
? CD ?3 e Q. C «*—
e
f O F
JC f F F ?
O UL O co O
»?)
 f
E
£
til

CS
e
o
t/3
In
0)
Q.
OD 2
e
%o ?

«2
<U

F
SM« >
e (D
o U
U O

*/3t

CxS 0)
.4>
?:
O

~V4 Jl" t»3 B

0)

Ja '«Ä -·-"
_=* "*^ K"
il *~ Ü
S «· ·»
e -3 -3
S o
V
tu

a
?»
Wi

JD

St
CA
S
e«

O
O

CJ

0) e
13 > (D
> ?)

?
JC
CtS
(L)
•9
en
o
(?
E
?-
?
Q.
Q.
O
X
ra
O)
C
?
?*—

O
C

O LL o <? CC O
VO
 f
E
e
¦s
UJ

e
o
V)
l>
Oi
a
S
M a
?
?- o

? I
3
>>
> > S
a f
ID o ¦o
u o
2
ß
,O

<u

SZ
¦a
o
CD <?>
-ta»
s
a>
? SZ ¦tal
X
<u
a>
JS
g- -5> -

a O
V
(? Z
?
?. °°
a
V
U

CD a»

C^ »?
**- -4"
?.
co
¡*
O VJ
B
03
as
a»

< ca CD
O
O
X!
U

e«

> (D e
-3 ?
t 3 (U
?
1= ffl E e
(0 a.
O ?
JC ? <u o
O LL o O
 UJ

a
o
U
a
N M) ?:
CD
a
% ?
o

I
?
JS*
^> F
S
D a
o
?
N ?
H +*
a
OO .o
M

a Cl, 9i
?
r* r ^ <**
?:
3
P Ö O
O V7
O
?
>> CO

V-I H a
Z ?
¦*-»

O Cw ?
Cu
J3
O
<
tí -**¦¦

? a
a»
CO
O

(-< -g -g <u
?.
a
a>
CO ?«

??
•o
-Aj
UJ
Q co
'¦¦J
U U
a

PQ
< as
?
?

a>

ce

f C
3 > F TD f
> F XJ
U '¦«3 ?-
«3 «? e Q. C
co f "c
S5 f
^* ? O
CO
F
Ct U.
O
O O O

00
 i
Ul

«
a
o
??
U
Q)
a.
W) JB
s
iE
M o
a
«S
J=

f
> S
> s
o
?
•s

D W
¦4-t
o
2
S
.O

tí S î
I S I
s 0
P-H a ¦"?H JC
-H» 0 a O
0 >> V)
g Si s
QJ
0
0>
9)
JS
H-J
¦3 P1
s o
g í> j> 2
G; »*? r™1 <?
T^ 00 u-
O.
Ü ?
U

Uh -W
»9
V
W "**

Q rt
JS
r t ?-
a?

C
< <0 O «
a»
J=
a>
O
O

et

> ? ?3 C
3 > (? F ?
F U CX X
a> E
?-
Q.
JZ 0) ? O 0) O
O ? LL (? a: O
OS
 F
E
Ití
tu

ß
O
w
?-
?
°> J»·
ß o»
•¦? ^y

>> a>

* > E?
§ I
V aj > ß

ta-
co
££:^
3 © C
L, <"-> ? O
O *
tö
fc> O
**-* ^
<? t.
¿i S

&« <u

o ? r^ ° <u
JS
¦w _^
"^ ^ o
»s
3
p
£3 £ T>>
S o
S z
**2 CT ?

a
V
CO In
G'** 1>- 11—
fr*
v£>
i^
di

e
to ce
4>

cm
O
O
o
ce

4>

(U C
Z5 > "O OJ
> co (U TJ
f
•ico U
E Q.
a.
X C
T3
1C
cu
JZ
O
CD
U-
S5 (U
u
O
O o
co
<u O
O
 f
E
Sì
H
tu

a
o
M
S.
N CU
a.
WO
C O)
X
X '%O
a
.o

CD

N > E
C ?
O TJ
H O
5
e/) C
«? ß
(U
£
o¦f> 2oj —-1%
O A *^? ¦?-
a ? Ä
JZ
-^. S g s?
O C/5 ? ? 3 O
in
»—I ? <?> <~>
B
,? <" 3
? ? ¿
"S^P
CLI

O ¿ àti S
^ -« >>
•S f>s r"
ans e
?—, o
? 1 C/} cu
?
<-< a» <i> CU
I*
Lj ? .'"? a
X ·-—· OO
tu

O <u

Ph ? ce
OO
W L.
CU
Q ?
C
U

C3 O O
O

cu

CJ
cu

> C
3 .> f cu
F t JD ö Q. ¦g
tu to E Q. C <4—
C
xz ? ? cu ? O O
O LL. O U. CO CC O
 f
E
ff
¦?
UJ

e«
a
o
M m
1«
<U
a.
>- wo
s
X X
O
ZS
,O

>
N
Z) ^1 03

>^ > 2
H X
a
o
?

O
U
CO C
F ?, F
> f > ,O
£¿ > ° J> .F
(? o +-.
O 3
-·—>
CL CO E 's ?:
o O
O O to
o
s—
>> co

s- **- nj a
c F a»
o
?. ^^ a»
Ol
o J=
o >,
c: -j^ TJ f CD

—>
E s O

O)
? f f kl
?? jc ?: Q.
QJ
I —. CO
O
*^
ce
u_ J3

LU ^_ UO ¡M

f ^r
Û S
X3 co
O O CN
CÛ O
O
< (Ü O J=

> TP C
Ji? > OJ
t ? ta
F t CL ? ¦g
a? CO
E a. C
C
JC O) ? o O
O LL U co or O
«?
 f
E
E
?
a?

a
N ?
(?
U
a>
Q. >»
OD ?
e o>
x
N %o
=3
«S
> ??

(U
cu
>
a (D
o T3
O
?¦ f * ¦**
s
5
4» — .O

.1U

* f « ?
W
C
M f ¦#

3f S a»

v-
C
•HI ^. ^ o

S N g
^ "^ fft ka
Q.

ri* 2 g f
-h C f>
W "" -r,
H ** ** a»
?>

? «I W ce
J3
-**

L.
cu

CA
e
" « « cu

y : . O
O

cu

e
3 > Xl (U
t 3 > ra (U ?
F -9 Ü Q. X
O) ro (TJ E ?. ro C
(U ?
JC ? (U O <¡>
o o
O O Ul in a: O
f«")
 f
E
£
'S
UJ

et
S
O
CA
k*
CU
a
m c
a
?

cu
X

CU
CU
>
C ?
O ?
U O
?
e
3 ,o
3

cu

a. a. g
E S UL. S
¡3 s ? O ,5?
----'US >> <?
SC „. ¡_¡
C
S
^
Sj
=
a CU
-w

CU
O . a
O)
J=

<
er re 5 <*>
_=«··" C
cu
"S
CA ?
cu
¡_
5J -= Q.
cu
U

X>
-*¦*
e«
J=
S-I
CU

e«
S
CTJ

O
O
J=
U
CU
!/1
«
cu

> C
3 ß)
?: 3 > >»
t ri O CL X C
CU CO E Ol cu C
x: cu tu O O F O
Ü LL O li. CO a: O
 F
E
g
UJ

?
U
a»
D.
©?)
a
?

«2

X "^
a>

> 2
s
?
U ?
-** 5
a
3» » s*
a» «2
S
ß S es
a? J= «~
ß ?. k_
EE-S
=3 ? E ¦S JC

—· O ? O
co
S » »
¿Mb
CO 53
S
a>
S« ß
a»

ft ?) ·*
tí ß ?»
-£« » <
^ JC O) ,5 ^i 2
ß3 ^^1 mjmi
?
"?
Z
» CD CD QJ

?.
?-

*** CD ^ ?)

ß xa CO
CD O CNJ
1.

CC TO O
e

?
?
J=

V
<?
C4

C
3 a >
t > co
? a. ? ¦o
OJ 1C E
co ?3 a. Ì5 e
?: ?
?-
S1 (U
?
U.
? o
O O
«?
 i
fi
s
tu

a
o
V)
?-
a»
Q.
M!
a
iE
o
N
ja
X V)
'S
> C
a f
H o ¦?
w O
*j 2
h a
X «2

»X.
?-, fc H
V) t e?
O 3
e >>
? I»· 'S
X
o en .£. V
o
?

'?- 3
(B
F* ?
a O
V) Z
<?
a
?

LA
Vl
a>
J2

-Xi
O
CNi

O V)
H-

?
V)
O
O
JS

<?
V)
ce

•e
.? F
>
C
(U
O a X ¦g
SI «5 E Q. C »?-

O LL
S1 (D
W o
U. or
OJ ?
O
O O
?
 f
E
S
'S
UJ

sa
e«
e
o
N Vl
U
a>
> a

X a
?
?

> QJ

N
> E
a
X o ¦a
W ?
? e
?
.o
oc >
> f >
° ? ?
O ? "d. -e a»
O- .S
? .2. Ü C s SZ
.2>
? X? F? gu ¦o
55
s- B

?. X!
O)
O)
?

?5
E ?- ? 5 s ?

X ¦-- ??
^_f f la
a

O
CO
s>
LJJ LO u

f ^t-
Q ¡?

TJ CO
O Ü CM
? ja t-
O
< CO ? O
J=
W
Ol
(?
«
Qi

F
3 > C
> (U

% •iro -9 "HS e
Q.
a. C
•s
?*
ja
SZ
O
(D 'ff « ?
O
? ?
(? ?:
e
?
O
 f
E
S
UJ

ß
O
te
kl

a
-?
N e
X

X
41

f
Si
>
3 G
O
f
•s
?
W
2
ß

CS o a %
«%
ib

s-
f go 3
? e» o O
?. o ^ o
o c Jj » C

e 3 ? o
o » * a>
a>
e ¿ g*.S J=
JS "O »

D M C ?
CO
*6* a>

a.

U. <* ¦? JS

?
¦?
•? ).

O Q e«
e
es ?
?
G JS

?
?
JS
U

es

3
JD
5 > f
C
OJ
?: O Q.
Iz -9 X
? CO re E
?—
Q. SS
C
f "e
JC (D 0) ? O ?
O LL O UL in (J
QC
 f
E
2
><
LU
N

X <u
Q.
?
N e s>
I
?

X
J=
(?
>-» cu

> >
S ?
a> 3
O
U O
5

Ili
3 O £

si
? ?
* o» o
C ? o
* E >»
3
O

a»
JZ
.3?

0 . o J=

-< e ? O W
S ?
* - ?

- ¦* * 3 ïa a ?-
a.

w & &
?
* — w
,e r·*
¦C G- G- 0)
-C
+J
C4
JS
+J

OJ

^ *- u> S

III ^ ^ W
cu

û -s « ja
eu
CA
?
?
J=

« ? t
< w o
? e
3 > ?
t 3 > as (U
.O ? Û. ? ?
f E CL e
a; ? 0) e
?: f ? ?
O O li- c/5 LC O
ON
 F
E

CU ai

s
o

Ol
a
u> ?
s
'?o ?

a
.o

¡?
f
cu

S
C-O O ?3
O
5

cu
5;
s ä
2?
«^
?:
s
?
Ü (?
B
CU
·*^
X
CU
??

<
re
a ?
cu

?-
a.
a>

«
J=
1.
CU
??
C/3
s
e«
CU

O
O
J=
CJ

cu

3 > TJ (U
.S3 > OJ TJ
a? O Q. X ¦g
cu co
E Cl m C
e
<u ?-
O (D
JC ? <u o
O a UL. IE O
O
 F
E
2
'S
LU
'¦?-»
e«
S
O
CA

Q-

S CO
X
'?O

N
Itu
JS
>
> C
a a?
E-H o
U ?
m 2
f f e

te è f ?> .©
b % f
O
g ß -° .?

O g F ?
? o

H
e; ?
o
B &· g.
i«!
2
«¦8 * -
? f
(d
s B
_ ¿*¡ ?

t«

•r-<
1

00 j2 f *Cf
C_t ?
f +^ ?-« 1«

O
Pl,
?
Q T3 CO
O
W ? m
<! «? o O
O

? e
>
3 > OJ
F t ? Q. X
E ?_ JS
C
?
JC O) (D O 0)
?- ?
O LL O CO a: O
fS
 5*
F
E
e
'S
tu

X a
?

ä
Vl

?
Q.
?
a
?
?

U
J=
3?
(D

> su S C
S
?
TJ
O
O

i «S a
e

ft<9,9
» F £ .4»
?«? « « JC

* « · O
?
• ? L
2 * S F
* 0 *
S
(U
+-*
X
a>
?

O
0)
?

F *· F A « 5! ?-
?.
Im

* f f Si
JC Jt ?
X «? I* ^ ^ 4*

0 ? f ?-
?)

Vl
a
«

J=

O
O
JS
U

pu ¿| ?* Z)
0)
0)
> C
cu
< 9 F s
3
Iz
CD
ja
s>
?
CO
•5
Q)
?3
e
Q.
Q-
O
0)
?
J3
TD
C
F
¦g
i*—
e
?: CL· a) 0) ?
O U- O CO ¡? O
 F
E
S
'S
ai

e
?
U
<?
a
©?
a
iE
?
a
a
J=

f
Ol
>
G f
O
'h— ? O
S

·> a £
a « "E f.
Q-. 3 Q E
O
O ? a ò ;?>
-**
W
s
a 5·· 3 ?
0 0 ' - ? JS
is #~
¦^ X) ti
¦?
*_
?» >
?-
a ?-7 ™ C O
QJ
(? Z
??* a»
ïî U. U ?-
a.
¦iz -C -C ?

J3
-W
es
JS
*j
>-
<?

e«
cu

O
O
JS
W

ce

cu e
=5 > X>
?: 3 f _> co ? f
.a o XJ
F t Q. ?
e
F CO ?»
TO TO E CL J5 f
?«—
C
JZ ? F O F
f ?» O
O LL. O C/3 O
 S
Vi 'S
UJ

C3
a
?
m
Ui
<u
a
N OJD
a CD
X
?

NI ?-)
(?

> ?
?- e
D a
?
U
f
¦s
?
H a
S
oo S "t? .?

^ O) <?>

f
.«
?:
•—^ ? ? ? .?»
0 er ? ? ?
a
C « 3
O fe ? ? X

tí ¿ ^s
'3?? fr 5
?? 3
i-j (?
a
QJ
Vl
QO S¿ a> a> Ul
l-M ¦ ?—I a.

O
Ph 00 m

Pi CO u

? a

<?
?
a>
m
O
O

e«

f
3 > C
F
to
? a
to E Q.
JT
O
CU
LL
SM £ a)
».
Ü
?
U.
O
W
?
LL
?
O
«?
 ro
o.
¦g en cr> CD O
ti CN en
(TJ ? ? O II
Q. cm ro ?
V.) KJ
en
<

in
 C
CO
a.
?

co JS
a to
?
f

«?
 O
O
C
?)
a.
? "03
? a?
a. -G: a?
CTS Q)
W =
O O
? > O
a>
?™ F CTS
CTl SZ O
F s? 'e
Q X ..K
O
UfJ f
f tr- f
JC F
en ai e» f
U Oi CD
X O
*?d "5 Q
"35 ?
e
JIT ?
to
? 'e
f
X H- (D ¿ ?
[TS G"**? (''"*! ?'*~y
"**?
O V^ '-·' '*-.' W ?--·

Tj
LLI

?-
Appendix C Name abbreviation of typeface
Typeface Name Abbreviation

Cooper Black CB
Berlin Sans FB BSF

Bernard MT Condensed BMC

Garamond Ga
Belwe Lt BT BLB

Playbill Pb

Harry Potter HP

Centaur Cr
Poor Richard PR

Jokerman Jm
Times New Roman TNR

Arial Al

Broadway Bw
Kino MT KM

Impact IP
Chiller Cl

Helvetica Ht

Bauhaus 93 Bh93

Kabel Kb

Onyx Ox

Rockwell Rw

Snap ITC SITC

Harrington Hr

Footlight MT Light FL

128
Appendix D Name abbreviation of
Personality Trait

Personality Trait Abbreviation

Cheerful Ch
Fearful Ff

Legible Lg
Attractive At

Creative Cr
Formal Fm

Sloppy Sp
Relaxed Rx

Friendly Fd
Confident Cn

129
Appendix E
Pearson's Correlation Coefficients
a»

C
O

(U

?3
a>

JS
"öS
ai

a.

U
s
?
fcK

a>

¡~
O
U
lu

s
O

a»

Oí
<n

a»
a.

s-
3
O
U-
£>
a

Í
«M
O

"C

e
o

4>
!m

?
IJ
?
'So

rt
É
Cl
B
O
Im

¦ß
a»

a*
CSS
U
c2
?
a

U
3
O

ß
??

?
???
O
#*
"E

ß
O

?-
?
U
CU

U
Cl
S-

s
o
CA

tu

t3
cu

Oí
CA
CU
U

o
ta

ß
CU

"S

C
O

0)
¡M
?-
?
?
a»
>

e
o
?

CL1

"ß

<?
a.

3
O
U-

?
O
• PN

Im

B
O

4»
?-
?
U
B
la
O
tu

?
£
ce
s
O
?

Dh

?3

U
Va
Ol
a

s
O
ta

Va
O
JJ

S
O

la
la
O
D.
Q.
_©

?
ë

e
o
¡?

(U

Ol

¡?
OS
OI
U
«S
<u
a

U
s
o
ta-

c<-
O

'C

B
O

(U
!m
?-
?
U
¦a

"3
oí

CS

Oh

.s
?
a

s
?
fe

?
O

-*¦*

S
O

O
S

e
o

O4

tu
a
P
U
3
O
Uh
£>
S
a»

?
?

JS
?«
?.
?
 e
-O

B
O
U

?
£

O
(?
U
cu

cu
QA
CA
CU
CU
.«
*¦*
CU
a

?«
s
?
[*
£>
B
CU

¿5
4M
O
*

?-
?
U
Appendix F
Rotated Component Matrix
Cheerful

Rotated Component Matrix

Component

TNR Ch .905

FL Ch .859

Cr Ch .852

Ga Ch .756

Rw Ch .682

Ht Ch .654

Al Ch .622

SITC Ch .773

Hr Ch .669

Bh93 Ch .640

Cl Ch .753

HP Ch .730

Jm Ch .545 .623

Kb Ch

BLB Ch .458

141
Fearful

Rotated Component Matrix

Component

BLB Ff .800

Ga Ff .790

TNR Ff .728

Cr Ff .722

Ht Ff .685 .454

FL Ff .634

SITC Ff .715

Al Ff .499 .695

Kb Ff .675

Rw Ff .586 .599

Bh93 Ff .467

Hr Ff .416

Cl Ff

HP Ff

Jm Ff

1 42
Legible

Rotated Component Matrix

Component

AlJ-g .857

Ht_Lg .794

Rw_Lg .743

Kb_Lg .731 .445

TNR_Lg .715

FL_Lg .549 .474

SITC_Lg .799

Hr_Lg .762

Bh93_Lg .751

Jm_Lg .736

HP_Lg .665

CI_Lg .658

GaJ-g
Cr_Lg .412

BLB_Lg

I43
Attractive

Rotated Component Matrix

Component

Cr At .837

TNR At .816

Ga At .785

Rw At .766

FL At .653

Ht At .648

Al At .618

Hr At .770

Cl At .767

HP At .756

Jm At .678

SITC At .614

BLB At .487 .503

Kb At

Bh93 At

144
Creative

Rotated Component Matrix

Component

Cr Cr .812

TNR Cr .803

Ga Cr .761

Rw Cr .723

Ht Cr .722

FL Cr .705

Al Cr .603

BLB Cr .544 .462

Bh93 Cr .833

SITC Cr .642

Kb Cr .641

Hr Cr .603

Jm Cr .531

Cl Cr

HP Cr

145
Formal

Rotated Component Matrix

Component

Ht Fm .813

Al Fm .803

Rw Fm .780

FL Fm .705

Kb Fm .684

TNR Fm .588

SITC Fm .842

HP Fm .835

Cl Fm .773

Jm Fm .751

Bh93 Fm .612

Ga Fm .866

Cr Fm .840

Hr Fm

BLB Fm

146
Sloppy

Rotated Component Matrix3

Component

TNR_Sp .798

BLB-Sp .770

PL_Sp .687

Ga_Sp .648

Rw_Sp .803

Bh93_Sp .670

AI_Sp .479 .630

Ht_Sp .552 .575

Cr_Sp .411 .563

Kb_Sp .498 .523

SITC-Sp .808

CI_Sp .794

Hr_Sp .732

Jm_Sp .675

HP_Sp

147
Relaxed

Rotated Component Matrix3

Component

Cr Rx .812

TNR Rx .795

Ga Rx .740

FL Rx .718

BLB Rx .696

Kb Rx .840

Bh93 Rx .620 .457

Ht Rx .512 .611

Al Rx .510 .563

Rw Rx .486

Cl Rx .777

Jm Rx .739

HP Rx .605

Hr Rx .437 .473

SITC Rx

148
Friendly

Rotated Component Matrix8

Component

TNR Fd .861

Cr Fd .816

Ht Fd .816

Ga Fd .757

FL Fd .730

Al Fd .721 .413

BLB Fd .434 .427

Kb Fd .686

SITC Fd .629

Rw Fd .479 .495

Jm Fd .749

Hr Fd .631

HP Fd

Cl Fd .556

Bh93 Fd

149
Confident

Rotated Component Matrix3

Component

Ga Cn .779

Ht Cn .765

TNR Cn .749

Cr Cn .742

Al Cn .692 .403

FL Cn .674

Bh93 Cn .735

SITC Cn .733

Kb Cn .712

Rw Cn .538 .606

Hr Cn .420

HP Cn .754

Jm Cn .718

Cl Cn .522

BLB Cn .402 .441

150
Appendix G Fifteen Examined Typefaces

Directness
Centaur

graphique
Garamond

graphique
Times New Roman

graphique
Arial

graphique
Helvetica

graphique

151
Rockwell

graphique
Footlight MT Light

graphique

Gentil ness
Belwe Lt BT

graphique
Bauhaus 93

9fophique
Kable

graphique
.Cheerfulness
Jokerman

Snap ITC

gropMçve
Harrington

;raphiqü£

Fearfulness
Harry Potter

graphique
Chiller

grapjïtyue

153