See discussions, stats, and author profiles for this publication at: [Link]

net/publication/299621647

Adaptive comfort relations and comfort temperature ranges from a field study
in undergraduate laboratories

Conference Paper · April 2014

CITATIONS READS

4 200

2 authors:

Asit Mishra Maddali Ramgopal

University of Galway Indian Institute of Technology Kharagpur
58 PUBLICATIONS 1,263 CITATIONS 53 PUBLICATIONS 1,109 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Asit Mishra on 06 May 2016.

The user has requested enhancement of the downloaded file.

Proceedings of 8th Windsor Conference: Counting the Cost of Comfort in a changing
world Cumberland Lodge, Windsor, UK, 10-13 April 2014. London: Network for
Comfort and Energy Use in Buildings, [Link]

Adaptive comfort relations and comfort temperature ranges from a

field study in undergraduate laboratories

Asit Kumar Mishra and Maddali Ramgopal

Department of Mechanical Engineering, Indian Institute of Technology Kharagpur,

India – 721302

Abstract
To ascertain comfort levels and effectiveness of available adaptive opportunities for classrooms in the
hot-humid regions of India, a thermal comfort field study was conducted in an undergraduate
laboratory class in Kharagpur. The study, carried out between January and April 2013, had
participation from 121 students and yielded 338 responses. Analysis of the results showed that comfort
temperatures found in the field study had close resemblance to the predicted comfort temperatures
evaluated from certain existing standard adaptive comfort equations. This was in spite of the standards
having been developed using observations from studies that had occupants with distinctly lower
metabolic rates than encountered in the current study. This is ascribed to the level of acclimatization
among the subjects as well as the availability of more adaptive avenues/more flexibility, during
laboratory classes.

Keywords: adaptive comfort; laboratory classes; natural ventilation; hot-humid climate; adaptive
opportunities

1 Introduction
It is well accepted that thermal environment of classrooms has a significant impact on
the teaching-learning process (Auliciems, 1972; Mendell and Heath, 2005). For a
growing economy like India, where the number of educational institutions, as well as
enrolments in them, is rising fast (UGC, 2008), any compromise of learning
environments is unacceptable. Historically, Indian classrooms have been naturally
ventilated (NV) along with ample use of fans. With India’s energy deficit keeping as
high as 26% recently (IEA, 2011), a sudden shift to use of air-conditioning in all the
classrooms is unlikely and unsustainable.
Results from several recent field studies vouch for the ease with which Indians adapt
to their local climate (Deb and Ramachandraiah, 2010; Dhaka et al., 2013; Indraganti,
2010; Indraganti et al., 2013; Pellegrino et al., 2012). From sustainability point of
view, adaptive comfort standards would be an ideal choice for ensuring comfort in
Indian classrooms. The user guide to Energy Conservation Building Code (ECBC),
issued by Bureau of Energy Efficiency, Government of India, also mentions the
adaptive comfort model given in ASHRAE Standard 55-2004 as an optional method
for determining comfort in NV buildings (BEE, 2009). Current concerns for energy
efficiency and sustainability mean that comfort standards of future will need to push
the limits to save the proverbial extra penny rather than settle for narrow comfort
bands. If the long term acclimatization of inhabitants in tropical climates could
broaden the upper limits of comfort standards that would give building designers
some important leeway. At the same time, a review of several field studies on thermal
comfort found that researchers working in classrooms often observe classrooms to
have lesser number of adaptive opportunities and more constraints in their use
(Mishra and Ramgopal, 2013). To the best of our knowledge, the study done by
Pellegrino et al. (2012) was the sole existing field study in Indian classrooms. So, to
add to the body of research in this important area, a field study was conducted during
the regular semester schedule of an undergraduate laboratory class. The objectives of
this study were:
 Verify the suitability of using existing adaptive comfort equations for
predicting comfort levels in Indian NV classrooms
 Ascertain if students are able to effectively adapt to their surroundings using
the available opportunities
 Check if sustained metabolic rates that are slightly higher than the near
sedentary levels, given for current adaptive standards, compromise the
predictive power of such standards
A portion of the findings from this field study has been reported elsewhere (Mishra
and Ramgopal, 2014). The previous work reported regression neutral temperature and
temperature zones for thermal comfort using different criteria for assessing
acceptability. Results showed that the students adapted to their NV surroundings well
and a majority of them were comfortable over the range of 20 to 31 °C. An adaptive
comfort equation (ACE) was also given based on survey data and predicted comfort
temperatures from this ACE were compared to predictions from certain standard
adaptive comfort models. The current work delves further along this line by
comparing observed comfort temperatures of the survey population on all 12 survey
days with prediction from different adaptive comfort models. Statistical significance
of the deviations in predictions from different models is presented. Comfort
temperature ranges over a single survey day, as opposed to the entire survey duration,
are discussed. A brief analysis of adaptive opportunities available and their adequacy
in facing the surroundings is also given.
2 Methodology
The study was conducted from January 2013 to April 2013 in undergraduate
laboratories of the Indian Institute of Technology Kharagpur (IIT). Kharagpur has a
Tropical savannah type (Aw) climate. The period from January to April comes under
Spring semester schedule of the institute. This particular semester was chosen as it has
both the warmest (April) and the coolest (January) months during which regular
classes take place. This meant that a wide spectrum of adaptive behaviour among the
students could be observed and it could be determined if the broad range of
temperatures, faced over a relatively short duration of four months, adversely impacts
adaptive abilities. Surveys took place on all twelve class days of the laboratory course
during the semester. As is usual in the institute, laboratory classes were held during
the post noon session. Details of the study location, description of the buildings,
subjects, questionnaire and survey methodology employed can be found in the earlier
work (Mishra and Ramgopal, 2014). Brief outline of methodology is provided in
Table 1.
For the activity level during a typical laboratory class, a met rate of 1.6 was used, as
per the recommendations in ISO 7730-2005 (ISO, 2005). All subjects had been
residents of Kharagpur for at least two and a half years and all of them were Indian
nationals. Hence, they had a significant level of acclimatization to the local climate.
Table 1. Survey methodology.

Survey Subjects briefly introduced to structure and purpose of the survey before
questionnaire being asked to answer the questionnaire.
Questions:
 Thermal sensation on ASHRAE scale
 Thermal comfort on Bedford scale
 Thermal preference
 Acceptability of thermal environment
 Air velocity sensation
 Humidity sensation
Survey  Survey started 75 to 90 minutes after beginning of class
activities  Time taken to complete survey activities: 35 – 40 minutes
 Measurements of dry and wet bulb temperature and air velocity taken
at shoulder height at five places around student groups (~10 in
number) as they filled up questionnaire
 One measurement of globe temperature per student group
 No major variations in thermal conditions around different student
groups. Hence, further analysis done with averaged out values of
environmental parameters
 Outdoor temperature data taken from the in-campus meteorological
station run by the Department of Physics and Meteorology
 Clo value estimated by matching student ensembles with a set of
standard ensembles
Instruments Instrument Make Range Resolution Remarks
Sling Local 0 to 1 °F Mercury
psychrometer 120 °F thermometers
Globe constructed -10 to 1 °C Alcohol
thermometer 110 °C thermometer,
plastic globe
of 70 mm
diameter
Anemometer Lutron 0.1 to 0.1 m/s Vane-type
AM4201 30 m/s

2.1 A summary of observations

A short summary of certain important outdoor and indoor parameters recorded during
the survey days is given in Table 2. Minimum and maximum values for the mean
thermal sensation vote (MTSV) on the seven point ASHRAE scale are also given in
this table. Daily mean temperatures were calculated as an arithmetic mean of daily
maximum and minimum temperature. The running mean temperature (RMT) given in
the table is a seven day running mean which was evaluated in a method similar to that
used in EN15251 (Nicol and Humphreys, 2010). Total number of students taking the
survey on any day varied between 23 and 32.
In Table 2, top represents the operative temperature; pv represents the partial pressure
of water vapour in air; va represents the average air velocity; and APD represents the
actual percentage dissatisfied, i.e., voting "Not Acceptable" on the survey
questionnaire to the question regarding acceptability of indoor thermal conditions.
 Table 2. Summary of observations.

Parameter Maximum Minimum

Outdoors
Daily minimum 25 °C 9 °C
Daily maximum 42 °C 20 °C
RMT 32.8 °C 16.9 °C
Daily mean 33 °C 14.5 °C
Indoors
top 35 °C 22 °C
pv 3.04 kPa 1.06 kPa
va 0.6 m/s 0.05 m/s
Average clo 0.91 clo 0.44 clo
MTSV 1.74 -0.73
APD 59% 0%

3 Results and Analysis

Over the twelve days of survey, 342 responses were obtained from a group of 121
students. Of these responses, four had to be classified as 'invalid'. Such a classification
was done when the subject answered 'no change' to the question on thermal preference
and yet found the environment 'not acceptable' or when the subject voted for an
extreme on the ASHRAE thermal sensation scale and preferred even more of the same
sensation on the thermal preference scale. For further analysis, only the 338 responses
that were not invalidated were used. The package R (R Core Team, 2012) was used in
all the statistical analysis performed.
Comfort temperature on any particular survey day was calculated by using the MTSV
and operative temperature recorded on that day, as inputs for the Griffiths' formula. A
value of 0.5/°C was used for the slope in Griffiths' formula, following the
recommendations of Humphreys et al.(2013). Since globe and operative temperatures
were almost equal during all the survey days, Griffiths' formula was evaluated using
operative temperature rather than globe temperature. This allowed the relationship
between indoor comfort and outdoors to be expressed in terms of operative
temperature.
3.1 Comparison of comfort temperatures with existing models
One of the goals at the start had been to check how well existing adaptive comfort
models would be able to predict the comfort temperatures obtained from the field
study. Certain differences between model predictions and study results were expected
on two counts. One was the higher metabolic rate of the subjects; the other was long
term acclimatisation of the subjects to a hot-humid climate. As standard adaptive
models, the ACE given in EN15251 (Nicol and Humphreys, 2010) and ASHRAE
Standard 55 (ANSI/ASHRAE, 2010) are taken. As the slope suggested by Humphreys
et al. (2013) in Griffiths' equation was used, the results are checked against
predictions of the equation put forth in the same work as well. The model proposed by
Nguyen et al. (2012) for hot-humid climates of South East Asia is also taken for
comparison considering the geographical proximity of India with South East Asia and
certain shared cultural traits. A more recent model proposed by Toe and Kubota (2013)
for hot humid climates, which they have derived using a meta-analysis of the
ASHRAE RP-884 database, is another model considered in the comparisons.
Both ASHRAE Standard 55-2010 and EN15251 recommend adjustments to the
comfort temperature in presence of enhanced air velocity. So, these adjusted values
are also taken as two sets of predictions when comparing the field study data. With all
the fans being off during January and February, air velocities remained close to 0 and
thus comfort temperatures during these months were not adjusted. On the other days,
comfort temperatures from EN15251 prediction were adjusted using the day's
50
recorded average air velocity as an input to the formula: 7  (Nicol and
4  10 va
Humphreys, 2010). Since the average air velocity never exceeded 0.6 m/s, as a
correction, 1.2 °C was added to the ASHRAE Standard 55 predictions
(ANSI/ASHRAE, 2010). An overview of the different modes used for comparison is
given in Table 3.
Table 3. Overview of adaptive comfort models.

Comfort equation Outdoors Survey Correction for

metric population enhanced air
velocity
ASHRAE tc=0.31 tout + 17.38; PMOAT Worldwide 1.2 °C
Standard 55 R2 = 0.70 (for 0.3 m/s < va
< 0.6 m/s)
EN15251 tc=0.33 tout + 18.8; RMT European 50
R2 = 0.358 7
4  10 va
(for va > 0.1 m/s)
Nguyen et al. tc=0.341 tout + 18.83; Monthly mean Hot, humid n.a.
R2 = 0.52 temperature regions of
South-East
Asia
Humphreys et tc=0.53 tout + 13.8; RMT Worldwide n.a.
al. r = 0.89 (preferred)
Toe and tc=0.57 tout + 13.8; MOAT Hot, humid n.a.
Kubota R2 = 0.64 regions from
RP 884
database
For the comparisons, Δtc is defined as Δtc = tc – tc,p, where tc is the comfort
temperature from the field study and tc,p is the predicted comfort temperature from
different models. Values of Δtc on the twelve survey days for the aforementioned
different models are plotted in Figure 1. From Figure 1 it can be observed that both
the ASHRAE model and ASHRAE adjusted model consistently under predict comfort
temperatures of this field study. However, for both EN15251 and Standard 55 models,
adjustments due to enhanced wind velocity bring the predictions closer to the
measured values. It is also observed that the closest predictions come from Nguyen et
al.'s model, Toe and Kubota model, and the EN15251 enhanced air velocity adjusted
model.
Paired Wilcoxon signed rank test is used to check if differences found between
predicted and observed values are significant. A first observation is that the prediction
set from the adjusted models for ASHRAE and EN15251 are significantly different
from the original model's prediction set (p value of 0.01 and 0.02 respectively).
Among all the models considered, the models whose predictions are not significantly
different from the observed values, at a 5% significance level, are the adjusted
EN15251 model, Toe and Kubota model, and Nguyen et al.'s model (p values of 0.30,
0.31, and 0.22 respectively).

Figure 1. Differences in comfort temperature found between study results and predictions of existing
models

This exercise yielded a few notable results. One was that use of adjustments for
enhanced air velocities is very useful for accurate comfort temperature prediction
especially when considering locations like India where fans are ubiquitous in NV
buildings. Secondly, the overall good performance of Toe and Kubota and Nguyen et
al.'s model showed that a model developed for similar climatic and cultural conditions
would be better at predicting comfort temperatures of occupants. But most
importantly, what is found is that in spite of a higher metabolic rate (more than 20%
higher than that of seated office occupants), the neutral temperatures found in this
study are not starkly different from the predictions of some of the existing ACEs.
With increased activity, one might have expected a drop in the comfort temperatures.
But the maximum difference found on any day is with the ASHRAE Standard 55
model where the model under predicts comfort temperature by 5 °C. This could be
due to the level of acclimatization of the occupants along with other contributing
factors. This is further discussed in Section 3.4.
3.2 Formulating an adaptive comfort relation with outdoors
As remarked upon in the previous section, different existing ACEs tend to use
different indices for the outdoors. So, before trying to relate indoor comfort conditions
with outdoors, a check was done to see if the different outdoor indices that might be
used as input to the model are significantly different from each other. Once more
paired Wilcoxon signed rank test is used and it is found that at a significance level of
5%, values for PMOAT, MOAT, RMT, and current month's mean temperature are not
significantly different. Also, values of PMOAT calculated using 9, 11, and 13 day
averages – instead of seven days – are not significantly different from the seven day
PMOAT.
Regression equations are developed between the indoor comfort temperature and the
following outdoor temperature indices: RMT, PMOAT-7 day, PMOAT-9 day,
PMOAT-11 day, PMOAT-13 day, MOAT, and the current month's mean temperature.
All these regression relations were found to be significant at 0.1% level. The R2
values of these relations are given in Figure 2. R2 values for the relations using RMT,
PMOAT-7 day, and MOAT are not very different though the value is highest when
using MOAT – similar to what Toe and Kubota (2013) found in their analysis. On the
other side though, use of more number of days in calculating PMOAT reduces
strength of relation quite rapidly. As any relation involving MOAT would require
knowledge of the current day's temperature, such a relation would have limited utility
as a dynamic/real-time prediction tool. In Equation 1, the regression relation between
PMOAT-7 day (tPMOAT) and tc is given, and in Equation 2, the regression relation
between MOAT (tMOAT) and tc is given.

tc  0.53tPMOAT  15.23; R 2  0.924, p  0.001 (1)

tc  0.49tMOAT  15.45; R 2  0.928, p  0.001 (2)

As already observed though, use of Toe and Kubota's model, or Nguyen et al.'s model,
or the EN15251, with adjustment for higher air velocity, gives results that are
statistically indistinct from the results that the above equations would provide. Thus,
Equations 1 or 2 are not proposed as yet another ACE in the already existing
multitude of ACEs. These equations are provided here to be informative rather than
normative.

Figure 2. R2 values of correlations using different indices of outdoor temperature

3.3 Temperature ranges for ensuring comfort

The results had shown that over the entire survey duration, 80% occupants were
comfortable over a temperature range of 20 to 31 °C (Mishra and Ramgopal, 2014).
This however represents the effects of an entire range of adaptive abilities occupants
employed as the season changed from winter to summer. For example, over the
survey duration, minimal clothing value observed was 0.39 clo while maximal was
0.99 clo. Over a single day or in particular during the three hour duration of the class,
full spectrum of adaptive opportunities can rarely be brought into play. It was required
to see what range of change in operative temperature during class timings on a day
would still leave 80% of occupants comfortable. To this end, the method used by
Nicol and Humphreys (2007) of correlating the deviation between tc and top during
survey period (Δtn = tc – top) with the percentage of occupants who found their thermal
environment to be acceptable (PS) on that day is utilised. Equation 3 gives a second
order polynomial fit between Δtn and PS.

PS  3.6tn2  3.22tn  96.38; R 2  0.91, p  0.001 (3)

Coefficient of the first power of Δtn in Equation 3 failed to achieve significance at 5%

level (p=0.15). So, ignoring contribution of the first-order term, Equation 3 is
rewritten as:

PS  3.6tn2  96.38; R 2  0.91, p  0.001 (4)

Equation 4 thus shows that around 4% people would be dissatisfied even if there is a
perfect match between tc and top. It also shows that corresponding to an 80% value for
PS, Δtn value needs to be ≈ ± 2 °C. This value is quite similar to the value obtained by
Nicol and Humphreys (2007) from their analysis of the SCATS database. However,
this finding comes with a qualifier. The current studies were all limited to post noon
periods. So, the temperature ranges experienced by the occupants were not typical of
the whole day's but rather of the hottest part of the day. Thus, the swing of 2 °C that is
found to be okay with 80% of the subjects may not be extrapolated to express the
temperature swings allowable over an entire day. Rather, such a range is
representative of the comfort ranges for the warmer half of a day. The comfort
temperatures found during the survey days along with the ACE line for PMOAT and
allowed temperature drifts are presented in Figure 3.

Figure 3. Comfort temperatures and adaptive comfort equation

3.4 Adequacy of adaptive opportunities in laboratories

Activity levels and patterns for laboratory and lecture classes are markedly different.
Apart from the higher activity level, laboratory classes are also longer. For example,
in IIT, while average lecture classes are of one hour, average laboratory classes are
three hours long. On the positive side though, students have a lot less restriction on
their adaptive behaviours. They do not have to occupy specific seats and maintain a
consistent posture. They are free to move about and break the monotony. During
summer months, students were observed to take multiple breaks for drinking water
(drinking water being provided within the lab premises). They clustered around or
under fans while not actively engaged with their respective experimental set ups.
Changes made to clothing were frequent. During winter, depending on the conditions,
students take off or zip up their jackets, sweat-shirts. In summer, they tend to loosen
the first couple of buttons on the shirt/t-shirt or fold up the sleeves of full sleeve shirts.
As discussed in Section 3.1, the higher metabolic rate did not cause a lowering of
comfort temperatures as given from certain accepted adaptive comfort models. The
availability of more adaptive opportunities in settings of the current study, compared
to those available in lecture classes or offices, is believed to be the reason behind this.
As discussed by Baker and Standeven (1996), availability of adaptive opportunities
can ameliorate the stress produced by a thermal stimulus. While this thermal stimulus
could be in terms of temperature, it could also be in terms of metabolic activity. Thus,
an enhancement in available adaptive opportunities is found to be able to successfully
deal with slightly increased met rates and currently existing adaptive comfort models
are able to predict comfort conditions without being grossly erroneous. It is however
likely that at a certain point increased metabolic rate will be high enough that
enhanced adaptive opportunities will not be effective. Finding where this tipping point
occurs would require further studies.
4 Conclusion
Results from the current field study validated the use of EN15251 model, with
adjustment for higher air velocity, Toe and Kubota’s model, and Nguyen et al.'s
model for predicting comfort conditions in NV classrooms of the hot-humid regions
of India. This emphasizes on the use of adjustments for air velocity when using an
ACE for locations where use of fans is widespread as well as highlights the fact that
an ACE developed from data collected in similar climatic conditions can often be
more useful than a global model.
In agreement with ASHRAE Standards 55 recommendations, PMOAT calculated
over seven days was found to be a suitable index for outdoor conditions. And in what
can be termed as a further validation of the principles of adaptive comfort, greater
adaptive opportunities were observed to successfully counteract slightly higher
metabolic rates without any appreciable changes in comfort zones.
Acknowledgements
The cooperation extended to us by our subjects is acknowledged. We also deeply
appreciate the help and support from laboratory staff members: Mr. R. Dey and Mr. P.
Roychowdhury. For giving us access to historical and current temperature data, we
are thankful to Prof. B. K. Mathur of the Department of Physics and Meteorology.
Our thanks also go to Prof. P. K. J. Mohapatra, Department of Industrial Engineering,
for the insights provided by him on statistical methods and analysis.

Abbreviations
ACE adaptive comfort equation
APD actual percentage dissatisfied
MOAT mean outdoor air temperature (daily mean)
MTSV mean thermal sensation vote
NV naturally ventilated
PMOAT prevailing mean outdoor air temperature
RMT running mean temperature
References
ANSI/ASHRAE, 2010. Standard 55-2010. Thermal comfort conditions for human
occupancy. ASHRAE, Atlanta.
Auliciems, A., 1972. Classroom performance as a function of thermal comfort.
International Journal of Biometeorology, 16(3), pp 233-246.
Baker, N. and Standeven, M. , 1996. Thermal comfort for free-running buildings.
Energy and Buildings, 23(3), pp 175-182.
BEE. 2009. Energy Conservation Building Code user guide. Bureau of Energy
Efficiency. New Delhi, India.
Deb, C. and Ramachandraiah, A., 2010. Evaluation of thermal comfort in a rail
terminal location in India. Building and Environment, 45(11), pp 2571-2580.
Dhaka, S., Mathur, J. , Wagnerb, A. , Agrawal, G. D. and Garg V., 2013. Evaluation
of thermal environmental conditions and thermal perception at naturally ventilated
hostels of undergraduate students in composite climate. Building and Environment, 66,
pp 42-53.
Humphreys, M. A., Rijal, H. B., and Nicol, J. F., 2013. Updating the adaptive relation
between climate and comfort indoors; new insights and an extended database.
Building and Environment, 63, pp. 40-55.
IEA, 2011. Energy balances of non-OECD countries. International Energy Agency.
Paris, France.
Indraganti, M., 2010. Using the adaptive model of thermal comfort for obtaining
indoor neutral temperature: Findings from a field study in Hyderabad, India. Building
and Environment, 45(3), pp 519-536.
Indraganti, M., Ooka, R., and Rijal, H. B., 2013. Field investigation of comfort
temperature in Indian office buildings: A case of Chennai and Hyderabad. Building
and Environment, 65, pp 195-214.
ISO, 2005. ISO 7730: Ergonomics of the thermal environment Analytical
determination and interpretation of thermal comfort using calculation of the PMV and
PPD indices and local thermal comfort criteria. 3rd edn., Geneva.
Mendell, M. J. and Heath, G. A., 2005. Do indoor pollutants and thermal conditions in
schools in influence student performance? A critical review of the literature. Indoor
Air,15(1), pp 27-52.
Mishra, A. K. and Ramgopal, M., 2013. Field studies on human thermal comfort – An
overview. Building and Environment, 64, pp 94-106.
Mishra, A. K. and Ramgopal, M., 2014. Thermal comfort in undergraduate
laboratories – A field study in Kharagpur, India. Building and Environment, 71, pp
223-232.
Nguyen, A. T., Singh, M. K., and Reiter, S., 2012. An adaptive thermal comfort
model for hot humid South-East Asia. Building and Environment, 56, pp 291-300.
Nicol, F. and Humphreys, M., 2007. Maximum temperatures in European office
buildings to avoid heat discomfort. Solar Energy, 81(3), pp 295-304.
 Nicol, F. and Humphreys, M., 2010. Derivation of the adaptive equations for thermal
comfort in free running buildings in European Standard EN15251. Building and
Environment, 45(1), pp 11-17.
Pellegrino, M., Simonetti, M., and Fournier, L., 2012. A field survey in Calcutta.
Architectural issues, thermal comfort and adaptive mechanisms in hot humid climates.
Proceedings of 7th Windsor Conference: The changing context of comfort in an
unpredictable world, NCEUB, Windsor, UK.
R Core Team, 2012, R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. URL: [Link]
[Link]
Toe, D. H. C. and Kubota, T., 2013. Development of an adaptive thermal comfort
equation for naturally ventilated buildings in hot-humid climates using ASHRAE RP-
884 database, Frontiers of Architectural Research, 2(3), pp 278-291.
UGC (2008), Higher education in India – Issues related to Expansion, Inclusiveness,
Quality and Finance. Secretary, University Grants Commission, New Delhi.

View publication stats