ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
The Survey Method in Architectural Research
Purpose of the unit
To introduce the use of Survey strategies in architecture and the broader built
environment and illustrate its application in both academic research and practice.
Expected learning outcome(s) The student should understand of basic terms,
characteristics, data collection and analysis techniques as used in architecture.
Mode Of delivery
Lecture and case studies
Unit content
Definitions. Characteristics. Standards of Survey Research: reliability, validity,
representativity, response rate, statistical significance. Data Collection Techniques.
Measurement scales: Nominal, Ordinal, Interval and Ratio. Data Collection Tools & Tactics:
Questionnaire, Observation, Sorting, Archives, Mapping. Use of statistics in analysis in survey
strategy. Mapping in architectural surveys: urban analyses, micro-scales spaces, space
syntax, mapping in Nairobi Housing studies. Strengths and Weakness of Survey Strategy.
Further Reading.
1
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
1 Introduction
Survey strategies 1 in architecture aim to relate spatial design parameters to function in
order improve its utility and other desired goals of the architect. They rely on fairly large
numbers — large 'n' [unlike Case Studies that use singular; or small numbers — small 'n'].
The basic strength of the strategy its capacity to use a representative sample and thus mostly
gauges positions of the entire population. It is also based Deductive reasoning mostly realised
through statistical analyses. Though it uses fairly large numbers, it employs limited numbers
of variables. This facilitates easier comparisons and correlations. The common tactic for
surveys is the Questionnaire. [as shall be shown other tactics can be used in architecture].
2 Definition of Basic Terms
Survey strategies rely on established methodological structures and employ common
terminologies. Some are hereby defined and explained.
i) Population
Sometimes called the term 'population' refers to the entire group of
individuals, events or objects a researcher wishes to investigate. Other than
individuals architects may observe events, buildings and processes — that thus
constitute the population. One must define the population that they want to
generalize results beforehand. One must list and define all the units that could be in
used for the study. In architecture can be defined by the characteristics that are likely
to affect space and design including humans and non-humans! Hence it could be
people, houses, buildings, building….. This can be further characterized by
geographical delimitations i.e. neighbourhood, city, and even country amongst
others.
ii) Sample
Given the size of the population, it may demand time and resources to investigate all of
them. One identifies a representation called a sample. This is carefully selected to reflect
characteristics of the population. Each member of this group is referred as a unit of analysis,
if human this could be called a 'respondent' or 'interviewee'. As already mentioned sampling
is not haphazard and demands systematic structure and indeed Sampling Theory 2 has been
further elaborated especially in the Social Sciences. Variable
These are the characteristics or attributes measured amongst the units of analysis and thus
assume different values based on circumstances. In surveys these are based on real world
1
In Other literature it is called Correlational Research [see: (Croat, Correlational Research, 2002)]
elements, and even processes.
2
Hence we encounter terms like: simple random sampling, cluster sampling, intentional sampling, self-
selection, casual interviewing and quota sampling.
2
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
and are naturally-occurring [and not constructed as in laboratories / workshops used in
experiments].
iii) Dependent variable
A Dependent variable is variable that is influenced.
v) Independent variable
Independent Variable (s) is one or more variables that influence the dependent variable.
vi) Data
This refers to all information gathered in research. It includes Primary data captured from
empirical work i.e. from the sample. Secondary data is gathered from other sources like:
articles, books, and established sources.
vii) Statistics
This is a system of science for organizing, describing and analysing quantitative data.
Descriptive statics only describes data. Inferential statistics draw inferences about a given
phenomenon based on the randomly used sample. The purpose of inferential statistics is to
test hypotheses and facilitate the researcher to generalize findings from the sample.
3 Characteristics of Survey strategy
The key characteristics of surveys include: (i) the focus on naturally occurring pattern, (ii) A
focus on measurement of specific variables, and (iii) use of statistics in the analysis,
A. Focus on naturally occurring patterns
In architecture, this takes the form socio-physical relationships i.e. social use of designed
space. The strategy thus assumes the researcher only measures variables of interest and
analyses their interrelationships. The aim is to redesign the space or guide future designs.
This characteristic is shared with Case Study strategy.
B. Focus on measurement of specific variables
A researcher aims to remain focussed by Identifying specific measureable parameters. Our
focus of space and functionality means there must be the spatial element and human use.
Often we measure how an existing design affects usability and social interactions. Surveys
are thus firmly rooted in the Quantitative paradigm.
This theory is not an objective in this Unit as it is rarely used most architectural research.
C. Use of statistics to clarify relationships
These range from pure descriptions [descriptive statistics] of frequency to correlations and
even describe the strengths of the relations in regression analyses. The depth a researcher
depends on the nature and aims of the study. Statistics are also used to test hypotheses.
Computer software that use statistics include: SPSS and Excel. Caution is however necessary
as statistics can distort reality and in fact such disconcertingly titled books like 'How to Lie
with Statistics' by Huff (1973)!
3
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
4 Standards of a good survey research
The judgement of a good research quality and usefulness that is used mostly in the
questionnaire surveys is based of five basic factors: Reliability, Validity, Representativity,
Response Rate and Statistical Significance.
This is the exactness in all phases of the research, theoretically so that another project
carried out the same way would have the same result. It thus concerned with internal
properties of measure that may contaminate scientifically acceptable outcomes.
The relevance of thee data to the research question, the appropriateness of the questions
as measuring the qualities they are intended to measure. It thus refers to relationship
between the data and the variables being measured.
C. Representativity
All units in the population [or universe] have known probability of being chosen in the
sampling procedure. A representative sample is important if one wants to generalize the
results of the research. A random sample does NOT mean haphazard as sampling is done
systematically to ensure representativity.
D. Response rate
In social science-related survey [sometimes necessary in architectural schemes like Housing
and Urban Design], the percentage of the sample who responded to questionnaire is called
the Response Rate and is important for validity and reliability.
E. Statistical significance
The size [in relations to the universe] of a selected sample definitely informs on the quality
and thus generalizability of the results. This measure is called Statistical Significance and has
been deeply investigated in statistical research literature. This is important if we opt for a
representative sample only and in available software [SPSSI this determined as one of the
outputs. If, on the other hand, one studies the ENTIRE population, this is NOT necessary.
5 Data collection tactics in architecture
Although Questionnaires as a commonly used in surveys and this is thus frequently confused
as the only survey tool, others also exist. These include Observations, Mapping, Sorting and
Archives, discussed here:
A. Survey Questionnaires
Questionnaires as structured set Of questions with standardised responses developed by
the researcher based on their synthesis of the problem. The main advantage of the survey
questionnaire is the broad scope of information that it can capture. The information
includes: demographic characteristics, behavioural habits, opinions and attitudes. Further a
fairly large segment Of the population, the sample, can be assessed in a fairly short time.
This however comes at the expense of in-depth understanding Of the issues. For depth one
4
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
employs qualitative analyses possible in Case Studies. Questionnaire must be developed
based on well-described rationale that emanates from the research questions or hypothesis.
Figure 1 is a structured question in a Questionnaire segment used from study in the US.
B. Observation
Observations are necessary tactics for understanding design and gaining useful knowledge
for improvements. For observations to be verifiable one needs use appropriate tools and
develop appropriate tactics for their use. These include time-lapse video filming and other
photography from cameras. These tools ought to be authenticated and well-described in
the methodology sections Of the work.
Sorting is tactic used in both research and practice for unearthing user-perceptions of
design. Here, respondents are asked sort various options [pictures, words) in order priority
[of say a design, function4 The objective is to identify the optimum design choices, values,
d. Archives
The use of archival data to guide includes sourcing records from an authority to
determine how an existing built-environment has shaped and influenced user-
reactions. For instance, Oscar Newman used police crime records to study 'defensible
spaces'.
He found out about
privacy gradients in housing, crime rates and building type and their location. About
crime, he found it increased with building, for instance.
5
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
F•ure
D. Mapping
Mapping is a handy tool that architects would value given their strengths in graphical
expression. This uses maps and Other graphical representations of a space by the architect
researcher to document intensity or otherwise of use at various scales of space ranging
from rooms to urban neighbourhoods.
6
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
7
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
6 Measurements
Surveys, especially employing the questionnaire, focus on measurement of specific
variables.
and the humm
Our focus of space and functionality means there must be the spatial
element and human use. Often we measure how an existing design affects
usability and social interactions.
In the example of Nairobi Open spaces the spatial parameters that
can easily be measured include (for instance): total area, amount
sittable space (benches), amount of green space (grass), amount of
paved space (Hard surface). These variables originate from the
researcher's observation, experience, literaturemas in deductive
theory. Since these variables are fixed for the particular natural
context, they are normally called: Independent Variables.
With these identified variables the may want relate them to the
performance of the space (s) under review. This can be measured
through various ways including popularity by observation [sheer
numbers at a particular time of the day...lunchtime?]. These
variables shall vary depending on the effectiveness of the design and
constitute the success or otherwise of the design. They thus called:
Dependent Variables. Other measures include seeking opinions
through questionnaires.
A researcher aims to remain focussed by identifying specific measureable parameters. The
Scales of Measurements in surveys include: Nominal, Ordinal, Interval and Ratio scales,
briefly described.
A. Nominal
These are discrete categories based on names and categories. [Like: Yes / No; male /
female...)
B. Ordinal
These ordered in order of value and ranking. [Like: A. strongly agree B. agree, C.
disagree D. strongly disagree] C. Interval
More precise measurement based the exact distance / interval between
measurements.
Here zero is missing or arbitrary [like a thermometer, is 360c twice as hot as
D. Ratio
This has higher precision with exact zero value and is reliant an ordinary number
system. [e.g. 20, is twice 101]
8
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
7 Use of statistics in analysis in survey strategy
These range from pure descriptions [descriptive statistics] of frequency to correlations and
even describe the strengths of the relations in regression analyses.
Descriptive statistics
These are used for describing the finding and assume various graphical forms including
piecharts, histograms... Figure 11 shows a small sample of four house types, the uses
extensions were used for using a horizontal bar-chart. Figures 12 & 13 illustrate the piechart
and the vertical bar-chart, of the many other options. The choice depends on many factors
determined by the researcher and nature of information to be conveyed.
9
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
Type sa
Type
5.2
Type
4.2
Type4.1
o 20 40 60 80 100 120
140 160
o commercial use • residential use 180
Proportion of Commercial uses in
Extensions
Type 5.2
Inferential Statistics and Correlations
Using statistical tools one can establish if an event influences another in design. In the
illustration four variables (QI-Q4] used to evaluate neighbourhood attachment in two
neighbourhoods is assed. Using statistical instruments of significance one can evaluate the
value each factor influences attachment. Statisticians decide on these figures for
acceptability (Here, a correlation significance level of 0.01 was preferred. SPSS usually auto
10
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
generates this data once inputs are keyed]. In the data, the ratings by the residents for both
rated the neighbourhoods positively. The type of statistics are called Inferential statistics
More complex studies aim to establish more bases of the statistical outcomes and often opt
for complex statistics called Regression Analyses. These are found more useful for more
advanced research unlikely for early researchers. This is found less likely useful in Applied
Research but some academic researchers in architecture may find it useful. Regression
analyses mostly test hypotheses. Thus they explain the reasons for these choices thus
support or falsify hypotheses.
8 Mapping in architectural Surveys
The following sections illustrate how mapping has been used in research in architecture
published in various books in other research.
A. in
Kevin Lynch's famous book, 'The Image Of the City' (1975), is classic example of the use of
maps derived from interviews that were utilized in evolving concepts urbanism in
American cities [Boston, Los Angeles & Jersey City). He went further to develop famous
concepts of urban perception: paths, nodes, landmarks, districts and edges.
11
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
Figure
P. 43)
In urban analyses mapping techniques are Often used to evaluate urban legibility
and hence design interventions in neighbourhoods, streets and even entire cities.
Bentley, etal„ (2005)'s book on urban design is useful in the use of mapping that
architects Often adopt. These establish the nature of both physical forms and
activity patterns.
12
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
micro-space level mapping is used variously to assess spatial parameters like furniture layout
to determine space-usage by dwellers in houses. John Zeisel (Inquiry by Design: Tools for
Environmental Behaviour Research, 2006, p. 167)has used in Environmental-Behaviour [E-Bl
studies he calls 'Physical Traces' [see: Fig. 61. Mapping of physical traces use such recording
tools like: annotated diagrams, drawings, photographs and counting.
Mapping studies
at micro-
scales
13
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
a p.
Hillier & Hanson (The Social Logic of Space, 1984)developed Space Syntax analysis in
mapping micro-level as well as urban space. This has been used in Housing Transformation
studies by (Housing Themselves-User-lnitiated Transformation of Government-Built Housing
in Developing Counties, 2000). Figure 7 illustrates graphically an example of how growth of
dwellings can understood and analysed [details not shown]. Computer algorithms now
employ the theory for urban analysis based on Hillier & Hanson.
14
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
D. Mapping Nairobi Housing Transformation analysis
A study, I undertook in Nairobi housing Dweller-initiated Transformation IDITs] studies used
mapping techniques whose Outcomes are illustrated in Figures 8 and 9 below. These show
the use annotated drawings at various scales of the Dwelling Unit and the Cluster as used in
Buru-Buru Estate-….
15
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
9 Strengths and weaknesses of the Survey Strategy — a summary
STRENGTHS WEAKNESSES
can clarify the researcher cannot control the experiments,
relationships among two or levels of degrees of variables quasiexperiments (can be
more naturally occurring manipulated)
variables
well suited for studying the less well suited for exploring Case Study [narrow-
breadth Of a setting or a the setting or phenomenon in butdeep]
phenomenon [broad- depth
butshallow
can establish predictive cannot establish causality experiments,
relationships quasiexperiments
o
Table 1: Stre f Strat
10 Further Reading
Bentley, Alcock, A., Murrain, P., McGtynn, P., & Smith, G. (2005). Responsive Environments -
a manualfor designers. Oxford: Elsevier.
Bryman, A., & Cramer, D. (2005). Quantitative Data Analysis with SPSS 12 and 13 - A Guide
for Social Scientists. London: Routledge.
Groat, L. (2002). Correlational Research. In L. Groat, & D. Wang, Architectural Research
Methods (pp. 173-202). Canada / USA: John Wiley & Sons.
Groat, L., & Wang, D. (2002). Architectural Research Methods. Canada / USA: John
Wiley &
Hillier, B., & Hanson, J. (1984). The Scrial Logic Of Space. Cambridge: Cambridge University
16
ARCHITECTURAL RESEARCH METHODS lecture 05 12/11/2024
Lynch, K. (1975). The Image of the City. Cambridge, Massachussets: MIT Press.
Makachia, P. A. (2012b). Transformation of Housing in Nairobi. Saarbrucken: LAP Lambert
Academic Publishing.
MO, L. (2004). Research Methods for Architects and Planners. NTNU, Town and Regional
Planning, Faculty of Architecture, Planning and Art. Trondheim: unpublished.
Newman, O. (1972). Defensible Space-People and Design in the Violent City. New York:
Macmillan.
Tipple, A. (2000). Housing Themselves-User-initiated Transformation of Government-Built
Housing in Developing Counties. Liverpool: Liverpool University Press.
Zeisel, J. (2006). Inquiry by Design: Tools for Environmental Behaviour Research
(revised edition ed.). Monterey: Brooks / Cole Publishing Company.
17