International Journal of Academic Research in Business and Social Sciences

Dec 2015, Vol. 5, No. 12

ISSN: 2222-6990

Assessing Infrastructural Development in Ogoniland in

the Niger Delta Region of Nigeria: A Participatory
Development Approach
Ojide Makuachukwu Gabriel*
Socioeconomics Unit, International Institute of Tropical Agriculture, Ibadan, Oyo State
Nigeria

Nwosu Emmanuel
Economics Department, Faculty of Social Sciences, University of Nigeria, Nsukka, Enugu State
Nigeria

Aramide Josiah O.
Physics Department, Faculty of Science, Adeyemi Federal University of Education, Ondo State
Nigeria

Ikpeze Nnaemeka I.
Economics Department, Faculty of Social Sciences, University of Nigeria, Nsukka, Enugu State
Nigeria

DOI: 10.6007/IJARBSS/v5-i12/1963 URL: [Link]

Abstract
Availability and access to basic infrastructure in rural areas across Nigeria has been identified by
many as a crucial component for national economic development. The impacts of
infrastructural facilities on quality of life and overall development cannot be underestimated. In
light of the above, this paper assesses infrastructural development in Ogoniland through a
participatory development approach. Using a multistage sampling method, four hundred
households were surveyed in the community. The ordinal logic regression model indicates that
infrastructural development in Ogoni community has been influenced by government
interventions and oil companies through their corporate social responsibility programmes.
Keywords: Infrastructure Development, Government, Oil Companies, Land Degradation, Oil
Spill

Introduction
In most developing economies, rural communities form the bedrock of economic development.
This is because these rural communities provide most of the basic agricultural and non-
agricultural products for economic development activities. Incidentally, these rural

354
[Link]

Electronic copy available at: [Link]

 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

communities have suffered neglect in terms of infrastructure development which has made
them less attractive to live (World Bank, 1994).
It is against this background that this paper examines infrastructural development in Ogoni
community (also known as Ogoniland) in the Niger Delta region of Nigeria. Apart from the need
for an in-depth and intensive study, the choice of Ogoniland for this paper is based on the fact
that the community was among the first places where oil was found in a commercial quantity in
Nigeria Shell began drilling in Ogoniland in 1958 (NEST, 1991 as cited in Ojide, 2015) .
Ogoni community consists of people of distinct indigenous minorities living in an area of 1,000
square kilometers on the south-eastern fringe of the Niger Delta Region of Nigeria. Given an
average population growth rate of 2.50 (2007 2010) and a population figure of 831,726 as
published by the National Bureau of Statistics (2206), the 2010 population of Ogoni people is
estimated to be around 914,899 (SAro-Wiwa, 1995; UNPO, 2009; World Bank, 2010).
Administratively, Ogoniland has four local government areas (LGAs). They are Eleme, Gokana,
Khana, and Tai with populations of 209,972; 251,711; 323,639; and 129,577 respectively (Ojide
& Ikpeze, 2015).
The concept of infrastructure has been viewed from different perspectives in recent times. In
general, however, infrastructure refers to those basic services and structures without which
primary, secondary, and tertiary productive activities cannot function. According to Khan
(1979), rural infrastructural facilities can be classified into three main types. They are physical
infrastructure (such as road, water, electricity, storage and processing facilities), social
infrastructure (such as health and educational facilities, community centres, and security
services), and institutional infrastructure (such as credit and financial institutions). This paper
focuses on three aspects of physical infrastructure, namely: access road, potable water and
electricity. FAO (2005) observed that rural infrastructure plays crucial role in poverty reduction
and overall socioeconomic development of households in rural communities of most
developing economies.
As an oil producing community, Ogoniland has been a rural area of interest to both the Nigerian
government and the oil companies in terms of intervention programmes aimed at
infrastructural development of the community. Many times, planning for such interventions
follow the top-down approach, as against the bottom-up or participatory approach, without
considering inputs or perceptions of knowledgeable local nationals. This paper, therefore,
examines the intervention activities of government and oil companies aimed at improving
infrastructure in Ogoni community. The study follows a participatory development approach
and for this purpose, we adopt the definition of participatory development concerned with
peoples perception of themselves and their challenges, interests and needs (UNDP, 2006;
OECD, 1995; Ojide, 2015). One fundamental assumption we make is that if locals are sufficiently
aware of infrastructure impacts in their environments, then they should effectively participate
in infrastructural development in their communities. Thus, the null hypotheses tested in this
paper are:

H0: Government interventions in Ogoniland do not have significant impact on basic

infrastructure in Ogoniland

355
[Link]

Electronic copy available at: [Link]

 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

H0: Oil companies interventions in Ogoniland do not have significant impact on basic
infrastructure in Ogoniland

Methodology
The data used in this research was obtained during a household survey. A multistage sampling
method was employed. Ogoni community was purposefully selected. A sample size was
obtained for the community. Then, using a quota method, the sample size was proportionally
distributed among the local government areas (LGAs). Enumeration areas were randomly
selected in each LGA. Finally, households were randomly selected in the chosen enumeration
areas within each LGA. Interview schedule was adopted as the survey instrument. The
sampling units were households in Ogoni community. Using 2010 estimate, distribution of
households in Ogoni community shows that Eleme had 45,397; Gokana had 54,422; Khana had
69,973; while Tia had 28,015 (National Bureau of Statistics, 2006; World Bank, 2010; Ojide &
Ikpeze, 2015).
The sample size formula specified by Yamane (1967), was applied (see equation 1):
N
s ...................1
1 N (e 2 )
s = required sample size.
N = the population size.
e = the degree of accuracy expressed as a proportion (.05).
Using the above formula, 400 households were obtained to make up the sample size. This
sample size was distributed in proportion to number of households in each local government
area in Ogoniland. That is, 92 households, 110 households, 141 households and 57 households
sampled from Eleme, Gokana, Khana and Tai respectively.
Structured interview schedule was predominately closed-ended questions to enhance response
rate and easy merging of data from all the four local government areas. The questions were
developed based on reviewed literature and preliminary interviews. Some multiple choice
questions also allowed respondents to comment further where necessary. As a result of the
sensitive nature of this survey, indigenes of Ogoni community were used as enumerators. They
were trained on general techniques for successful questionnaire administration. In addition,
they were given detailed review of each question why the question is asked and the expected
range of responses including how to ask the question to avoid leading question bias. Also,
they were instructed to adequately explain to the respondents the purpose of the survey as to
avoid, as much as possible, biased responses. The use of educated indigenes of Ogoni
community enhanced communication and reduced security risks given the emotional and
political nature of the subject of interest and the study area. The questions were asked by the
enumerators who filled-in the responses into the interview schedule. This reduced the chances
of misinterpreting the questions. The respondents in each of the selected households were the
heads of the households or their representative (who must be a spouse or adult son/daughter,
where an adult is a person not less than 18 years old).

356
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

Pilot Stage and Test of the Instrument

Research experts (including an indigene of Ogoni community) conducted the face and content
validation of the interview schedule. The reliability of the instrument was determined during
the pilot study of 30 households randomly selected in Tai local government area (which is one
of the local government areas of the study). The interview instrument was administered to the
30 households. Their responses were tested using Split-half reliability index coefficient alpha
(Cronbach, 1951). In split-half technique, the coefficient alpha is calculated using equation 2
(Allen & Yen, 1979):
N
..................2
( N 1)[1
Var (Y ) ]
i

Var ( X )
where N = number of items
Var(Yi) = sum of item variances
Var(X) composite variance
On the average, the research instrument achieved about 83% reliability.

Model Specification
The impacts of government and oil companies interventions on basic infrastructure in
Ogoniland were evaluated using ordinal logit model as stated in equation 3 (Brenton, 2010).
y* k xk k .............3
where y* is an unobserved, continuous, underlying tendency behind the observed ordinal
response (rating). The Xk represent the independent variables, while the k represent the
associated parameters. The error term (k) captures stochastic (unobserved) variation. It is
assumed to be distributed logistically.
Relating the unobserved y* to Y through a series of cut points, is as represented in equation
4:
Y 1 if y* 1
Y 2 if 1 y* 2

.......4
...
Y j if j 1 y *

where Y is the rating and the s represent thresholds of y* that delineate the categories of the
ordered response variable. These threshold parameters are restricted to be positive where
each one is greater than the previous. The first parameter 1 is normalized to 0 so that one less
parameter has to be estimated. That is not a problem because the scale of the latent variable is
arbitrary (Borooah, 2001; Ojide & Ikpeze, 2015).
To avoid confusion and misinterpretation of estimates, Y is restricted to a five-point
Likert item or less measuring impacts of oil companies interventions and government
interventions on basic infrastructure in Ogoni community. Using equation 3, equation 5 was
estimated. Variables are as defined in table 1.
357
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

Road f(GIrd , CSRrd , LD)

water f (GIw, CSRw, OS , income).......5
Electricity f (GIel, CS Re l )

Table 1: Definition of Variables

Variable Description
Code
Electricity Electricity supply (very low=1, low=2, mild=3, high=4, very high=5)
LD Land Degradation (very low=1, low=2, mild=3, high=4, very high=5)
OS Oil Spillage (very low=1, low=2, mild=3, high=4, very high=5)
Road Accessible roads (very low=1, low=2, mild=3, high=4, very high=5)
Water Portable water (very low=1, low=2, mild=3, high=4, very high=5)
Household income (18000 & Below=1, 18100 - 50000=2, 50100 - 100000=3,
Income
100100 250000=4, > 250,000=5)
CCGG to GIw below were coded as: low=1, average=2, high=3, very high=4
CSRel Corporate Social Responsibility of oil firms towards supply of electricity
Corporate Social Responsibility of oil firms towards provision of accessible
CSRrd
roads in the community
CSRw Corporate Social Responsibility of oil firms towards provision of portable water
GIel Government intervention towards supply of electricity
Government intervention towards provision of accessible roads in the
Gird
community
GIw Government intervention towards provision of portable water

Results and Discussion

Household survey was conducted in all the four local government areas (LGAs) in Ogoni
community. The survey started in December 3, 2013 and ended in January 17, 2014. Complete
responses from sample size of 400, as specified, were successfully collected.
Analyses of categories of respondents in all the four local government areas of Ogoni
community (pooled data) indicate that greater proportion (54%) of the respondents was
household head. The 46% non-household head respondents were wives (22.8%), sons (34.8%)
and daughters (42.4).
Majority (81.5%) of the households surveyed were indigenes of Ogoniland while the rest were
non-indigenes residing in Ogoni community up to the time of the survey (table 6). About 79.5%
of the households had lived in the community beyond 10 years, 13.5% had lived in the
community between 5 year to 10 years; and 7% had lived in the community for a time period
below 5 years.
358
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

Breakdown of literacy level of household heads reveals that only about 3.8% of the household
heads in the community had no formal education. Majority of the household heads were
literate with primary school (10.3%), secondary school (26.3%), national diploma (22.8) and first
degree/post graduate degree (37%). The results of the estimated ordinal logit models are
presented in table 2.
The three models estimated are presented in table 2.

Table 2: Ordinal Logistic Analysis of Economic Impact Models

Road Water Electricity
Variable Value 1, 2, 3, 4, 5 1, 2, 3, 4, 5 1, 2, 3, 4, 5
Predictor Coef Odds Coef Odds Coef Odds
Ratio Ratio Ratio
Const (1) 1.6447* 0.6308 0.9807*
{0.000} {0.074} {0.000}
Const (2) 3.1997* 2.0058* 2.4207*
{0.000} {0.000} {0.000}
Const (3) 4.9785* 3.7543* 4.0203*
{0.000} {0.000} {0.000}
Const (4) 7.1824* 5.9022* 5.4006*
{0.000} {0.000} {0.000}
GIw 0.7424* 0.48*
{0.000}
GIel 0.9784* 0.38*
{0.000}
Gird 0.9888* 0.37*
{0.000}
CSRw 0.4794* 0.62*
{0.002}
CSRel 0.4612* 0.63*
{0.000}
CSRrd -0.3913* 0.68*
{0.001}
OS 0.0885 1.09
{0.255}
LD -0.15383 0.86
{0.062}
Income 0.08895 1.09
{0.225}
Test that all slopes are 139.920* 61.152* 163.437*
zero (G) {0.000} {0.000} {0.000}
Goodness-of-Fit Test 280.986* 614.573* 131.546*
(2) {0.004} {0.000} {0.000}
359
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

Cases used 384 (96%) 382 (95.5%) 383 (95.75%)

Cases with missing 16 (4%) 18 (4.5%) 17 (4.25%)
values

A. Impact on Road
i. Overall Model
In this model, 96 percent of the observations were used while the rest were excluded due
to missing values. The goodness-of-fit test, Chi-square (2 = 280.986) with p-value (0.004),
indicates that the model is appropriate for the data. Similarly, the overall relationship
between the independent variables and the dependent variable is significant. This is
because the statistic G (139.92), with p-value of 0.000, indicates that there is sufficient
evidence to conclude that at least one of the estimated coefficients in the model is different
from zero. Thus, the independent variables are simultaneously significant.

The model examined three factors namely government intervention towards provision of
accessible roads (GIrd), corporate social responsibility of oil firms towards provision of
accessible roads (CSRrd) and land degradation (LD). The p-values of the predictors indicate
that for 0.05 alpha-level, there is sufficient evidence to conclude that Ogoni peoples
perceptions about government intervention towards provision of accessible roads (GIrd)
and corporate social responsibility of oil firms towards provision of accessible roads (CSRrd)
are statistically significant. However, unlike government intervention towards provision of
accessible roads, corporate social responsibility of oil firms towards provision of accessible
roads (CSRrd) is negatively associated with availability of access road in the community. On
the other hand, the result indicates that land degradation (LD) is not a significant factor
affecting access road in the community. In general, therefore, the null hypotheses about
government and oil companies interventions were rejected. However, while concluding
that government intervention has positive impact on availability of accessible roads, it is
observed that people of Ogoni community perceive oil company activities to have adverse
effect on access roads in the community.

ii. Cumulative Predicted Probabilities for each Score Category and Probabilities for the
Individual Scores of the Dependent Variable at the means of the Independent
Variables

Keeping the estimated parameters fixed (that is = 0), the cumulative predicted
probabilities for each of the five categories and probabilities for the individual scores of
access road in the community were calculated (table 3).

360
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

Table 3: Cumulative Predicted Probabilities of Availability of Access Road

Predictor Coeff Score Cum Prob Prob (individual
(score) score)

Const (1) 1.6447 1 0.838173 0.838173

Const (2) 3.1997 1 or 2 0.960823 0.12265

Const (3) 4.9785 1 or 2 or 3 0.993163 0.03234

7.1824 1 or 2 or 3 or 4 0.999241 0.006078

Cumulative scores (5) 1 or 2 or 3 or 4 or 5 1 0.000759

Table 3 indicates that Ogoni community has greater probability (0.838173) of being at the
lowest category in terms of availability of access road in the community. The regression result
suggests that land degradation is not a significant contributor to such low category of access
road. The analysis suggests that Ogoni peoples perception is that oil firms activities have
adversely affected availability of accessible road in the community. This is unlike government
intervention towards provision of accessible roads (GIrd) which shows positive impact on
availability of accessible road in the community.

B. Impact on Potable Water

i. Overall Model
In this model, 95.5 percent of the observations were used while the rest were excluded due
to missing values. The goodness-of-fit test, Chi-square (2 = 614.573) with p-value (0.000),
indicates that the model is appropriate for the data. Similarly, the overall relationship
between the independent variables and the dependent variable is significant. This is
because the statistic G (61.152), with p-value of 0.000, indicates that there is sufficient
evidence to conclude that at least one of the estimated coefficients in the model is different
from zero. Thus, the independent variables are simultaneously significant.

The model examined four factors namely government intervention towards provision of
potable water (GIw), corporate social responsibility of oil firms towards provision of potable
water (CSRw), oil spill (OS) and household income. The p-values of the predictors indicate
that for 0.05 alpha-level, there is sufficient evidence to conclude that government
intervention towards provision of potable water (GIw) and corporate social responsibility of
oil firms towards provision of potable water (CSRw) are statistically significant. Both
variables are positively associated with availability of potable water in the community. On
the other hand, the result indicates that oil spill (OS) and household income are not
significant factors affecting availability of potable water in the community. Therefore, the
null hypotheses about government and oil companies interventions were rejected with the
361
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

conclusion that government and oil companies interventions towards provision of potable
water in Ogoni community have positive and significant impact on effect on availability of
potable water in the community.

ii. Cumulative Predicted Probabilities for each Score Category and Probabilities for the
Individual Scores of the Dependent Variable at the means of the Independent
Variables

Keeping the estimated parameters fixed (that is = 0), the cumulative predicted
probabilities for each of the five categories and probabilities for the individual scores of
potable water in the community were calculated (table 4).

Table 4: Cumulative Predicted Probabilities of availability of potable water

Predictor Coeff Score Cum Prob Prob
(score) (individual score)
Const (1) 0.6308 1 0.652671 0.652671
Const (2) 2.0058 1 or 2 0.881405 0.229734
Const (3) 3.7543 1 or 2 or 3 0.977119 0.095714
Const (4) 5.9022 1 or 2 or 3 or 4 0.997274 0.020155
Cumulative scores (5) 1 or 2 or 3 or 4 or 5 1 0.002726

Table 4 indicates that Ogoni community has greater probability (0.652671) of being at the
lowest category in terms of availability of potable water in the community. The regression
result, however, suggests that oil spill is not a significant contributor to such low category of
availability of potable water in the community. On the other hand, the analysis result suggests
that government (GIw) and corporate social responsibility of oil firms towards provision of
potable water (CSRw) have contributed significantly to availability of potable water in the
community.

C. Impact on Electricity
i. Overall Model
In this model, 95.7 percent of the observations were used while the rest were excluded due
to missing values. The goodness-of-fit test, Chi-square (2 = 131.546) with p-value (0.000),
indicates that the model is appropriate for the data. Similarly, the overall relationship
between the independent variables and the dependent variable is significant. This is
because the statistic G (163.437), with p-value of 0.000, indicates that there is sufficient
evidence to conclude that at least one of the estimated coefficients in the model is different
from zero. Thus, the independent variables are simultaneously significant.

The model examined two factors namely government intervention towards provision of
electricity (GIel) and corporate social responsibility of oil firms towards provision of

362
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

electricity (CSRel). The p-values of the predictors indicate that for 0.05 alpha-level, there is
sufficient evidence to suggest that government intervention towards provision of electricity
(GIel) and corporate social responsibility of oil firms towards provision of electricity (CSRel)
are statistically significant. Both variables are positively associated with availability of
electricity in the community. Therefore, the null hypotheses about government and oil
companies interventions were rejected with the conclusion that government and oil
companies interventions towards provision of electricity in Ogoni community have positive
and significant impact on efficient electricity supply in the community.

ii. Cumulative Predicted Probabilities for each Score Category and Probabilities for the
Individual Scores of the Dependent Variable at the means of the Independent Variables

Keeping the estimated parameters fixed (that is = 0), the cumulative predicted
probabilities for each of the five categories and probabilities for the individual scores of
electricity in the community were calculated (table 5).

Table 5: Cumulative Predicted Probabilities of availability of electricity

Predictor Coeff Score Cum Prob Prob
(score) (individual
score)
Const (1) 0.9807 1 0.727247 0.727247
Const (2) 2.4207 1 or 2 0.918392 0.191145
Const (3) 4.0203 1 or 2 or 3 0.982369 0.063977
Const (4) 5.4006 1 or 2 or 3 or 4 0.995506 0.013138
Cumulative scores (5) 1 or 2 or 3 or 4 or 5 1 0.004494

Table 5 indicates that Ogoni community has greater probability (0.727247) of being at the
lowest category in terms of availability of electricity in the community. The regression result
suggests that government intervention towards provision of electricity (GIel) and corporate
social responsibility of oil firms towards provision of electricity (CSRel) have contributed
positively to level of availability of electricity supply in the community.

Conclusion

Infrastructural development can be influenced by government intervention and corporate

organizations through their corporate social responsibility programmes. Following the
perception of Ogoni people, this paper concludes that availability of potable water and
electricity supply in the community are significantly and positively influenced by government
363
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

interventions and by the corporate social responsibility functions of oil firms. Government
intervention also positively impacts the availability of accessible roads in the community.
Given their high literacy levels and the correlations of these with impact perceptions,
household heads prove that they can contribute meaningfully to infrastructural development
planning in their communities. Therefore, developmental and multilateral agencies willing to
improve infrastructure outlook in Ogoniland should consider engaging the household heads
during the formation of such development programmes to ascertain community priorities.

Acknowledgement

The authors wish to acknowledge the contributions of the following people in providing
suggestions, vetting the research instrument and proofreading:

Prof. Stall Madueme, University of Nigeria, Nsukka

Asst. Prof. Onyukwu Onyukwu, University of Nigeria, Nsukka
Dr. Augustus Legborsi An indigene of Ogoniland, River State, Nigeria

Corresponding Author
Ojide Makuachukwu Gabriel*
Socioeconomics Unit, International Institute of Tropical Agriculture, Ibadan, Oyo State, Nigeria
Email: embracegod1@[Link]

References
Allen, M.J. & Yen W.M. (1979). Introduction to measurement theory. Monterey, CA:
Brooks/Cole.
Borooah, V.K. (2001). Logit and probit: Ordered and multinomial models. Sage University
Papers Series on Quantitative Applications in the Social Sciences: 7-138, Thousand Oaks,
CA: Sage
Brenton, J. D. (2010). Massachusetts Landowner Participation in Forest Management Programs
for Carbon Sequestration: An Ordered Logit Analysis of Ratings Data. Master of Science
Thesis (Department of Resource Economics) Submitted to the Graduate School of the
University of Massachusetts Amherst.
Cronbach, L.J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika,
16(3): 297-334. [Accessed 10 October 2013
([Link] )]
FAO (2005). Plan of action for peoples participation in rural development: Twenty-sixth
session, FAO conference. Rome, 9-28, Nov.
Khan, A. (1979): Social Policy and Social Services. 2nd Ed. New York; Random House.
National Bureau of Statistics (2006). Federal Republic of Nigeria 2006 population census.
[Accessed 15 January 2013
([Link]/nbsapps/Connections/[Link])]
364
[Link]
 International Journal of Academic Research in Business and Social Sciences
Dec 2015, Vol. 5, No. 12
ISSN: 2222-6990

NEST (1991). Nigerias threatened environment: A national profile. Ibadan, Nigeria. Nigerian
Environmental Study / Action Team: 229
Ojide, M. G. (2015). Socioeconomic Baseline Study of Ogoniland in The Niger Delta Region Of
Nigeria. Researchjournalis Journal of Economics. 3(1): 8-15. [Accessed 20 July 2015
([Link]
Ojide, M. G. & Ikpeze N. I. (2015). Education, Health and Housing in Ogoni Community: Does
Government or Oil Firms Intervention Matter?. Asian Journal of Social Sciences and
Management Studies, 2015, 2(1): 44-52. [Accessed 14 Septembe 2015
([Link]
Saro-Wiwa, K. (1995). Complete statement by Ken Saro-Wiwa to Ogoni civil disturbances
tribunal. Roslindale, Assachusetts: Rat Haus Reality Press.
UNPO (2009). Ogoni. Retrieved from United Nation. [Accessed 20 June 2015
([Link]
World Bank (1994). Adjustment in Africa: Reforms, Results and the Road Ahead. A World Bank
Policy Research Report, Oxford University Press.
World Bank (2010). World Bank Databank. [Accessed 13 May 2015
([Link]
Yamane, T. (1967). Statistics: An introductory analysis, 2nd Ed, New York: Harper and Row, page
258

365
[Link]