Chapter Two: Sampling and Sample Designs

2.1. The Concept of Sampling

In most cases social research cannot be carried out on a social reality in its entirety; the
researcher usually has to settle for a sub-set of this reality. Thus sampling procedures are
often the first empirical operations in which the researcher must engage.

Sampling – that is to say, observing a part in order to glean information about the whole – is
an almost instinctive human act. We need only think of a cook tasting food to see whether
salt should be added, an examiner questioning a student on part of the examination syllabus,
or a journalist questioning a taxi driver to find out what ‘the people’ think of some topical
issue. Although common, however, the process of sampling was scientifically systematized
only in very recent times, with the introduction of the notions of chance and randomization.

Sampling is the procedure through which we pick out, from a set of units that make up the
object of study (the population), a limited number of cases (sample) chosen according to
criteria that enable the results obtained by studying the sample to be extrapolated to the whole
population. The mathematical formulations of sampling have been studied in depth, giving
rise to a specific branch of statistics known as sampling theory.

In survey research, the selected respondents constitute what is technically called a ‘sample’
and the selection process is called ‘sampling technique.’ The survey so conducted is known
as ‘sample survey’. Algebraically, let the population size be N and if a part of size n (which is
<N) of this population is selected according to some rule for studying some characteristic of
the population, the group consisting of these n units is known as ‘sample’. Researcher must
prepare a sample design for his study i.e., he must plan how a sample should be selected and
of what size such a sample would be.
Implications of a Sample Design
A sample design is a definite plan for obtaining a sample from a given population. It refers to
the technique or the procedure the researcher would adopt in selecting items for the sample.
Sample design may as well lay down the number of items to be included in the sample i.e.,
the size of the sample. Sample design is determined before data are collected. There are many
sample designs from which a researcher can choose. Some designs are relatively more precise
and easier to apply than others. Researcher must select/prepare a sample design which should
be reliable and appropriate for his research study.
2.2 Benefits of Sampling
As stated earlier, sampling has many advantages over census. Some of the advantages of
sampling are:
 a sample study usually is substantially less expensive than a census,
 sample usually produces information faster than a census does and has sampling
inquiry saves time,
 the result obtained by sampling often are almost as accurate and sometimes even
more accurate than those obtained from a census inquiry because in sample inquiry
the entire work is generally conducted by trained and experienced investigators,

1
  more detail information can be obtained from a sample survey than from a census
because a sample in many instances takes less time , is less costly and primarily more
care to be taken in census execution,
 Sampling errors can as well be estimated specially in case of random sample inquiry.
2.3. Some Basic concepts
Population
Population is the entire set of relevant units of analysis. It is the aggregate of all cases that
conforms to some designated set of specification. For instance, population can be defined as
all people reside in Ethiopia. Or a population can be defined as, by employing specification,
all students enrolled in Ethiopian universities. You can also define population as all
households reside in a given community, all farming households in your woreda, etc. A
population is the set of units that the sample is meant to represent. This is the general unit of
analysis of the research to which generalization is made from the sample.
The Sample Unit
A single member of a sampling population (e.g., a voter, a household, an event) is referred to
as a sampling unit. Usually, sampling units have numerous attributes, one or more of which
are relevant to the research problem. Sampling unit may be a geographical one such as state,
district, village, etc., or a construction unit such as house, flat, etc., or it may be a social unit
such as family, club, school, etc., or it may be an individual.
Sampling Frame
Sampling Frame is also known as source list. Once population is defined, the next step is
drawing a sample that adequately represents that population. A sampling frame is a list which
contains the names of the units of the population from which the sample is drawn. If the list
consists of the following qualities, then the sampling method becomes successful.
i. The list must be valid and up-to-date. Names of deceased persons must not be
included in the list.
ii. The list must be reliable.
iii. The source list /framework must be relevant to the study
iv. There should not be repetition in the list
v. The list should be exhaustive in nature. It must contain full information about the
units. The address of the units must be written in the list. For instance, suppose the
objective of the research is to see the achievement of sociology students (Regular,
Evening and Distance). The sampling frame for this can be a student roster.
An Observation Unit: An observation unit is also known as unit of Data collection is an
element or aggregate of elements from which information is collected.
Variable: Variable is a set of mutually exclusive characteristics such as sex, age,
employment status, and soon.
Statistic: A statistic is the summary description of a given variable in a survey sample. It is
statistical measures worked out on the basis of sample studies. A sample mean or a sample
standard deviation is an example of statistic. Statistical induction or statistical inference (the
movement from sample to the population) is made based on a statistic.
Parameter: A parameter is the summary description of a given variable in a population. All
the statistical measures based on all items of the populations are termed as parameters. A
population means or a population standard deviation is an example of parameters.
Finite and infinite populations: a population may be finite or infinite, depending on whether
the sampling units are finite or infinite. A finite population contains a countable

2
 number of sampling units, for example, all registered voters in a particular city in
a given year. An infinite population, on the other hand, consists of an endless
number of sampling units, such as unlimited number of coin tosses.
Sample: when the data serving as the basis for generalization is comprised of a subset of a
population, that subset is called a sample.
Sampling: is a statistical process in which we select and examine a sample from a general
population instead of considering the whole population.
Survey sampling: sampling is designed to produce information about particular
characteristics of a finite population is usually termed as survey sampling.
Sample size: is the total number of units of analysis in the sample.
Sampling error: is a mistake that was made in selecting a sample. However, it is,
statistically, a deviation from what actually exists in the population (Refer Sirkim,
1999: 199 - 200). Therefore, parameter value – statistic value = sample error.
Non sampling error: is an error in measurement, calculation, recording, etc.
Sample size determination:
The size of sample plays an important role in sampling.
The sample must be small enough to avoid unnecessary expenditure and large enough to
avoid sample error.
The size of the sample depends upon a number of factors which are stated below:
• Homogeneity or Heterogeneity: If the population of the universe is homogeneous, a
small size of the sample may serve the purpose.
• But if the universe is heterogeneous, then the sample must be larger in size.
• Law of statistical Regularity: This law indicates that the samples must be random
and representative of the whole universe.
• Only then, account results are obtained.
• Size of Questionnaire: The size of the questionnaire or schedule plays a vital role on
the size of the sample.
• If the questionnaire is small in size, the small sized sample can serve the purpose. But
if the schedule is large and the questions are difficult, the sample should be large in
size.
• Nature of sampling: If the cases are geographically scattered, small sample is more
suitable.
• If the selection of samples is done by the stratified sampling method, the reliability
and accuracy can be obtained by small sized samples.
• Taking the above criteria into account, various statisticians recommend different
sample determination formula

3
For instance, Sulvian has provided the following sample determination formula.
• The formula is written as n = N / (1 + Ne 2) where n = Number of samples, N = Total
population and e = Error tolerance.
• Example if our population N=48,227 and e= 0.05,
• n= 48,227
1+48,227(0.05)2
• n= 397
• If we add 10% non-response rate, the final n= 397+39.7= 437
2.4 Feature of Sampling
The most important features of sampling are:
the sample should be representative of the universe from where it has been taken,
there should remain no bias in selecting a sample,
it should be possible to measure or estimate the sampling errors, and
The results of the sample study in general should be applicable to all items of the
universe.

2.5. Types of Sampling

In general there are two methods of selecting a sample. These are
A. Probability sampling, and
B. Non probability sampling.
Probability Sampling
Probability sample designs permit the researcher to specify the probability (chance) of each
sampling unit of being included in the sample in a single draw from the population. In
probability sampling, the manner in which the sample units are selected is very important.
The process usually entails drawing subjects from an identified population such that every
unit in the larger population has precisely equal chance of being selected as every other unit
in that population. This process is called randomization (random sampling).
Probability Sampling Designs
From the above discussion, we have seen that probability sample designs permit the
researcher to specify the probability of each sampling units being included in the sample in a
single draw from the population. In such type of sample, there are several different designs
on the basis of various factors. Some of these designs include the following:
Simple random sampling,
Systematic random sampling,
Stratified sampling,
Cluster sampling,
Multistage sampling (stratified cluster sampling), and
Disproportionate sampling.
1. Simple Random Sampling

4
Random sampling may also be known as chance sampling or probability sampling. Under
random sampling each and every item of the universe has an equal chance of inclusion in the
sample. Individuals or elements included in the sample are not deliberately but by some
mechanical process. Here it is blind chance alone that determines whether the one unit or
another is selected. Simple random sampling is a procedure that gives each of the total
sample units of the population (noted by the letter “N”) an equal and known non zero
probability of being selected. For example, when you toss a perfect coin, the probability that
you will get a head or a tail is equal and known (50%) and each subsequent outcome is
independent of previous out comes.
Random sampling ensures the Law of Statistical Regularity which states that if on an average
the sample chosen is a random one then the sample will have the same composition and
characteristics as the Universe. This is the reason why random sampling is considered the
best technique of selecting a representative sample, a basic essential of a good sampling
technique.

Simple random sampling has the following advantages

▪ Gives each element in the population an equal probability of being selected into the
sample, and
Any one of the following methods is used to select samples randomly in simple random
sampling:
▪ Drawing lots or what is known as the lottery sampling. Under it we allocate to each unit
in the population a number; the numbered slips are then mixed thoroughly in a container
and withdrawn as a lottery either blindfolded or by rotating a drum or by any other similar
device.
▪ Arranging the units geographically, numerically or alphabetically and then selecting
every 10th or 15th (or choosing any other figure) item from the prepared list
indiscriminately.
The selection of any given sampling unit is random, that is, independent of the selection
procedure in this method and you can then estimate parameters with the confidence that they
are representative of the real values you would find in the total population. Random selection
procedures ensure that every sampling unit of the population has an equal and known
probability of being included in the sample; this probability is n/N where n stands for the
size of the sample and N for the size of the population. For example, let the population of
Ambo town (located in Oromia National regional state) is 42720 and a simple random sample
of 3230 is to be drawn, the probability of each sampling unit of the populations being
included in the sample is 3230/42720= 0. 075
Advantages and Disadvantages of Random Sampling
Advantages of Random Sampling
 The random sampling method is more representative since in this method , each unit
has equal chance to be selected,
 There is no scope for bias and prejudices,
 The method is very simple to use, and
 It is easy to find out the errors in this method.
Disadvantages of Random Sampling
 If the units of items are widely dispersed , the selection of samples becomes
impossible
 If the units or items are heterogeneous on nature or of different size and nature, the
random sampling method becomes inapplicable,
2. Systematic Sampling

5
Systematic sampling consists of selecting every K th sampling unit of the population after the
first sampling unit is selected at random from the total of sampling units. Under systematic
sampling only the first unit of the sample is selected at random and the remaining units are
selected at fixed intervals. The first selection is determined by same random process, such as
the use of a table of random digits. If you wish to select a sample of 10 from a population of
100 student of 2020/21 enrolled in sociology department of Ambo university, you would take
every tenth individual (K= N/n=100/10= 10). Suppose that the fourth student was selected;
the sample would then consist of individuals numbered 4, 14, 24, 34, 54, 64, 74, 84, and 94.

The basic assumption under this sampling technique is that all elements of the universe are
ordered in a manner representative of the total population. Systematic method is an easier
method of selecting a sample. It has the advantage of not requiring the use of random
numbers. In practice, when lists of the population are readily available but are of considerable
length, systematic sampling is generally used. Systematic sampling is more convenient than
simple random sampling when interviewer untrained in sampling techniques have to conduct
their sampling in the field; it is much simpler to instruct them to select every K th person from
a list than to have them use a table of random digits. Systematic samples are also more
amenable for use with very large populations or when large samples are to be selected. In
systematic sampling, each sampling unit in the population has a 1/K probability of being
included in the sample.

But there are certain inherent dangers in using systematic sampling. If there is hidden
periodicity in the population, systematic sampling is an inefficient method of sampling. There
is one danger involved in a systematic random sampling. The arrangement of elements in the
list can make a systematic sampling unwise. If the list of element is arranged in a cyclical
pattern that coincides with the sampling interval, it is possible that grossly biased sample may
be drawn. For example, a researcher's sampling frame is organized by married couples with
the male first and the female second. Such a pattern gives the researcher an unrepresentative
sample if systematic sampling is used. His or her systematic sample can be non-
representative and include only wives because of how the cases a reorganized. When his or
her sample frame is organized as couples, even-numbered sampling intervals result in
samples with all husbands or all wives.
Therefore, when we are interested in using systematic random sampling from a list, we
should carefully examine the nature of that list. If the elements are arranged in any particular
order, we should better use a simple random sampling.

Advantages
(a) This is a simple method of selecting a sample.
(b) It reduces the field cost.
(c) Inferential statistics may be used.
Disadvantages

6
(a) This is not free from error, since there is subjectivity due to different ways of systematic
list by different individuals. Knowledge of population is essential.
(b) Information of each individual is essential.
(c) This method can’t ensure the representativeness.
3. Stratified sampling: If a population from which a sample is to be drawn does not
constitute a homogeneous group, stratified sampling technique is generally applied in
order to obtain a representative sample. Under stratified sampling the population is
divided into several sub-populations that are individually more homogeneous than the
total population (the different sub-populations are called ‘strata’) and then we select items
from each stratum to constitute a sample. Since each stratum is more homogeneous than
the total population, we are able to get more precise estimates for each stratum and by
estimating more accurately each of the component parts, we get a better estimate of the
whole. In brief, stratified sampling results in more reliable and detailed information.
The following three questions are highly relevant in the context of stratified sampling:
(a) How to form strata?
(b) How should items be selected from each stratum?
(c) How many items be selected from each stratum or how to allocate the sample size of each
stratum?
Regarding the first question, we can say that the strata be formed on the basis of common
characteristic(s) of the items to be put in each stratum. This means that various strata be
formed in such a way as to ensure elements being most homogeneous within each stratum
and most heterogeneous between the different strata. Thus, strata are purposively formed and
are usually based on past experience and personal judgment of the researcher. One should
always remember that careful consideration of the relationship between the characteristics of
the population and the characteristics to be estimated are normally used to define the strata.
At times, pilot study may be conducted for determining a more appropriate and efficient
stratification plan. We can do so by taking small samples of equal size from each of the
proposed strata and then examining the variances within and among the possible
stratifications, we can decide an appropriate stratification plan for our inquiry.

In respect of the second question, we can say that the usual method, for selection of items for
the sample from each stratum, resorted to is that of simple random sampling. Systematic
sampling can be used if it is considered more appropriate in certain situations.

Regarding the third question, we usually follow the method of proportional allocation under
which the sizes of the samples from the different strata are kept proportional to the sizes of
the strata. That is, if Pi represents the proportion of population included in stratum i, and n
represents the total sample size, the number of elements selected from stratum i is n* Pi. To
illustrate it, let us suppose that we want a sample of size n = 30 to be drawn from a
population of size N = 8000 which is divided into three strata of size N1 = 4000, N2 = 2400
and N3 = 1600. Adopting proportional allocation, we shall get the sample sizes as under for
the different strata:
For strata with N1 = 4000, we have P1 = 4000/8000
and hence n1 = n . P1 = 30 (4000/8000) = 15
7
 Similarly, for strata with N2 = 2400, we have
n2 = n . P2 = 30 (2400/8000) = 9, and
for strata with N3 = 1600, we have
n3 = n . P3 = 30 (1600/8000) = 6.
Thus, using proportional allocation, the sample sizes for different strata are 15, 9 and 6
respectively which is in proportion to the sizes of the strata viz., 4000 : 2400 : 1600.
Proportional allocation is considered most efficient and an optimal design when the cost of
selecting an item is equal for each stratum, there is no difference in within-stratum variances,
and the purpose of sampling happens to be to estimate the population value of some
characteristic. But in case the purpose happens to be to compare the differences among the
strata, then equal sample selection from each stratum would be more efficient even if the
strata differ in sizes. In cases where strata differ not only in size but also in variability and it
is considered reasonable to take larger samples from the more variable strata and smaller
samples from the less variable strata, we can then account for both (differences in stratum
size and differences in stratum variability) by using disproportionate sampling design by
requiring: n1/N1 1 = n2 /N2 2 = ......... = nk /Nk 2k

where 1, 2 , ... and k denote the standard deviations of the k strata, N1, N2,…, Nk denote
the sizes of the k strata and n1, n2,…, nk denote the sample sizes of k strata. This is called
‘optimum allocation’ in the context of disproportionate sampling. The allocation in such a
situation results in the following formula for determining the sample sizes different strata:
ni= n*Ni i
N1 1 +N2 2 + ……+ Nk k for i = 1, 2, ….and k.
We may illustrate the use of this by an example.
Illustration 1
A population is divided into three strata so that N1 = 5000, N2 = 2000 and N3 = 3000.
Respective standard deviations are: 1 = 15, 2= 18 and 3 = 5.
How should a sample of size n = 84 be allocated to the three strata, if we want optimum
allocation using disproportionate sampling design?
Solution: Using the disproportionate sampling design for optimum allocation, the sample
sizes for different strata will be determined as under:

Sample size for strata with N1 = 5000

n1= 84 (5000) (15)/ (5000) (15) + (2000) (18) + (3000) (5)
=6300000/126000=50
Sample size for strata with N2 = 2000
n2= 84 (2000) (18)/ (5000) (15) + (2000) (18)+ (3000) (5)
= 3024000/126000 = 24
Sample size for strata with N3 = 3000
n3= 84 (3000) (5)/ (5000) (15) + (2000) (18) + (3000) (5)
= 1260000/126000 = 10
From what has been stated above in respect of stratified sampling, we can say that the sample
so constituted is the result of successive application of purposive (involved in stratification of
items) and random sampling methods. As such it is an example of mixed sampling. The

8
procedure wherein we first have stratification and then simple random sampling is known as
stratified random sampling.
LIMITATIONS
Stratified sampling needs reliable prior knowledge of the population in order to effectively
define the strata and allocate the sample sizes. The gains in the precision, or the reductions in
cost, depend on the quality of the information used to set up the stratified sampling design.
Any possible increases in precision are particularly dependent on strength of the correlation
of the auxiliary, stratification variable with the variable being observed in the study. Precision
may be reduced if optimal allocation is used and if the auxiliary variable used for the
optimization calculations does not accurately reflect the variability of observations for the
study.
Advantages
(a) It is (more precisely third way) a good representative of the population.
(b) It is an objective method of sampling.
(c) Observations can be used for inferential purpose.
Disadvantages
(a) Serious disadvantage of this method is that it is difficult for the researcher to decide the
relevant criterion for stratification.
(b) Only one criterion can be used for stratification, but it generally seems more than one
criterion relevant for stratification.
(c) It is costly and time consuming method.
(d) There is a risk in generalization.
As with simple random sampling, with a stratified sampling plan the investigator may
encounter difficulties identifying and gaining access to the sampled locations in the field.
Such limitations may reduce the expected gains in precision anticipated by using a stratified
sampling scheme.
(iii) Cluster sampling: If the total area of interest happens to be a big one, a convenient way
in which a sample can be taken is to divide the area into a number of smaller non-overlapping
areas and then to randomly select a number of these smaller areas (usually called clusters),
with the ultimate sample consisting of all (or samples of) units in these small areas or
clusters. Thus in cluster sampling the total population is divided into a number of relatively
small subdivisions which are themselves clusters of still smaller units and then some of these
clusters are randomly selected for inclusion in the overall sample. Suppose we want to
estimate the proportion of machine parts in an inventory which are defective. Also assume
that there are 20000 machine parts in the inventory at a given point of time, stored in 400
cases of 50 each. Now using a cluster sampling, we would consider the 400 cases as clusters
and randomly select ‘n’ cases and examine all the machine parts in each randomly selected
case.

Cluster sampling, no doubt, reduces cost by concentrating surveys in selected clusters. But
certainly it is less precise than random sampling. There is also not as much information in ‘ n’
observations within a cluster as there happens to be in ‘n’ randomly drawn observations.
Cluster sampling is used only because of the economic advantage it possesses; estimates
based on cluster samples are usually more reliable per unit cost.
(iv) Area sampling: If clusters happen to be some geographic subdivisions, in that case
cluster sampling is better known as area sampling. In other words, cluster designs, where the
primary sampling unit represents a cluster of units based on geographic area, are

9
distinguished as area sampling. The plus and minus points of cluster sampling are also
applicable to area sampling.
(v) Multi-stage sampling: Multi-stage sampling is a further development of the principle of
cluster sampling. Suppose we want to investigate the working efficiency of nationalized
banks in India and we want to take a sample of few banks for this purpose. The first stage is
to select large primary sampling unit such as states in a country. Then we may select certain
districts and interview all banks in the chosen districts. This would represent a two-stage
sampling design with the ultimate sampling units being clusters of districts.

If instead of taking a census of all banks within the selected districts, we select certain towns
and interview all banks in the chosen towns. This would represent a three-stage sampling
design. If instead of taking a census of all banks within the selected towns, we randomly
sample banks from each selected town, then it is a case of using a four-stage sampling plan. If
we select randomly at all stages, we will have what is known as ‘multi-stage random
sampling design’.

Ordinarily multi-stage sampling is applied in big inquires extending to a considerable large

geographical area, say, the entire country. There are two advantages of this sampling design
viz., (a) It is easier to administer than most single stage designs mainly because of the fact
that sampling frame under multi-stage sampling is developed in partial units. (b) A large
number of units can be sampled for a given cost under multistage sampling because of
sequential clustering, whereas this is not possible in most of the simple designs.
Nonprobability Sampling
Nonprobability sampling is applicable where the researcher has little initial control over the
choice of who is presented for selection, or where controlled selection of participants is not a
critical factor. Therefore, nonprobability sampling techniques are used for the following
reasons:
When probability sampling would be prohibitively expensive, and
When precise representativeness is not necessary.
The major designs utilizing nonprobability samples have been employed by social scientists:
 Accidental haphazard/Convenience/ sampling,
 Purposive sampling/judgment sampling,
 Quota sampling,
 Voluntary sampling,
 Snowball sampling,
 Event sampling,
 Time sampling.
A Accidental haphazard/Convenience/ Sampling: (sampling those most convenient; those
immediately available). Here, researchers may select
whatever sample units which are conveniently available. For example, a researcher may take
the first 200 people encountered on the street who are willing to be interviewed. In such type
of sampling, the researcher has no way of estimating the representativeness of the sample,
and therefore, cannot estimate the population parameter.
Limitations

10
  It does not allow the level of confidence/uncertainty of the researcher.
 It limits statistical inference
B Purposive Sampling/Judgment Sampling:
(enables you to use your judgement to choose people that are presented or are available that
best meet your objectives or your target groups). In the case of judgement sampling,
researchers select sampling units subjectively in an attempt to obtain a sample that appears to
be representative of the population. It totally depends on the researcher’s subjective
judgement which cannot be possible to determine why he/she judges why the sampling unit is
selected.
The main consideration of the researcher in employing this technique is the determination of
who can provide him with detailed information about the issue under the study. Therefore,
the researcher uses his own opinion to go to those people who have rich information and are
willing to provide him. Still the sampling frame and the representative ness of the
information collected is unknown. It is mainly applicable when it is used by an expert who
knows the population very well, to identify a typical case, to construct a historical reality, to
describe about an unfamiliar phenomenon etc…
C. Quota Sampling: the main goal of quota sampling is to select a sample that is as similar
as possible to the sample population. For example, a researcher
found that 48% of population was women. Therefore, in his
sample of 200 persons 96 were women.
D Voluntary Sampling :( the sample is self-selecting; they come forward voluntarily in
response to an appeal)
E ‘Snowball’ sampling: (is building up a sample through informants. You start with one
person – who then suggests another & so on)
F Event Sampling: (using the opportunity presented by a particular event, e.g. a conference,
to make contacts)
G Time Sampling: (recognizing that different times or days of the week or year maybe
significant and sampling at these times or days.

11
Chapter Three : Instrument design/ Questionnaire construction
Definition
A questionnaire is a form which is prepared and distributed for the purpose of securing
responses. Generally these questions are factual and designed for securing information about
certain conditions or practices, of which recipient is presumed to have knowledge. Goode and
Hatt have defined questionnaire as a device for securing answers to questions by using a form
which the respondent fills himself.
Questionnaires are essential tools as well as most directly associated with survey
research. Given its strong link to social survey, it is important to look at some of the
established techniques of questionnaire construction.
Design of the Questionnaire
In order to gain acceptance for his questionnaire, the researcher should design an appealing
format. Many unattractive questionnaires end up in a wastebasket rather in the hands of the
sender. To improve the attractiveness of the instrument, choose a title that is clear, concise,
and descriptive of the research project, and use well typed or printed questions that are
properly spaced and easy to read. It is generally advisable to group questions of a similar
nature together.
Principles of Good Question Writing
A good questionnaire forms an integrated whole. The researcher weaves questions together
so they flow smoothly. He or she includes introductory remarks and instructions for
clarification and measures each variable with one or more survey questions. Three principles
for effective survey questions are: Keep it clear, keep it simple, and keep the respondent's
perspective in mind. Good survey questions give the researcher valid and reliable measures.
They also help respondents feel that they understand the question and that their answers are
meaningful. Questions that do not mesh with a respondent's viewpoint or that respondent find
confusing are not good measures.

The principles of question writing are illustrated in the following 12 things to avoid when writing
survey questions. The list does not include every possible error, only the more frequent problems.

1. Avoid jargon, slang and abbreviation: avoid slang and jargon unless a specialized
population is being surveyed. Target the vocabulary and grammar to the respondent
sampled
2. Avoid ambiguity, confusion and vagueness: Ambiguity and vagueness plague most
question writers. A researcher might make implicit assumptions without thinking of
the respondents. For example, the question, "What is your income?" could mean
weekly, monthly, or annual; family or personal; before taxes or after taxes; for this
year or last year; from salary or from all sources. The confusion causes
inconsistencies in how different respondents assign meaning to and answer the
question. The researcher who wants before-tax annual family income for last year
must explicitly ask for it. Another source of ambiguity is the use of in definite words
or response categories. For example, an answer to the question, "Do you jog
regularly? Yes- No,” hinges on the meaning of the word regularity. Some respondents

12
 may define regularly as every day, others as once a week. To reduce respondent
confusion and get more information, be specific.
3. Avoid emotional language: Words have implicit connotative as well as explicit
denotative meanings. Words with strong emotional connotations can color how
respondents hear and answer survey questions. Use neutral language. Avoid words
with emotional" baggage" because respondents may react to the emotionally laden
words rather than to the issue. For example, the question, "What do you think about a
policy to pay murderous terrorists who threaten to steal the freedoms of peace-loving
people?” is full of emotional words.
4. Avoid prestige bias. Titles or positions in society (e.g., president, expert, etc.) carry
prestige or status. Issues linked to people with high social status can color how
respondents hear and answers survey questions. "Do you support the president's
policy regarding Kosovo?” will be answered by respondents who have never heard of
Kosovo on the basis of their view of the president.
5. Avoid double-barreled questions: Make each question about one and only one topic.
A double-barreled question consists of two or more questions joined together. It
makes a respondent's answer ambiguous. For example, if asked, "Does this company
have pension and health insurance benefits?" a respondent at a company with health
insurance benefits only might answer either yes or no. The response has an ambiguous
meaning, and the researcher cannot be certain of the respondent's intention.
6. Do not confuse beliefs with reality. Do not confuse what a respondent believes with
what you, the researcher, measures. A respondent may think that a relationship exists
between two variables but this is not an empirical measurement of variables in a
relationship. People's beliefs about a relationship among variables are distinct from an
actual empirical relationship.
7. Avoid leading questions. Make respondents feel that all responses are legitimate. Do
not let them become aware of an answer that the researcher wants. A leading (or
loaded) question is one that leads the respondent to choose one response over another
by its wording. There are many kinds of leading questions. For example, the question,
"You don't smoke, do you?" leads respondents to state that they do not smoke.
8. Avoid asking questions that are beyond respondents' capabilities. Asking something
that few respondents know frustrates respondents and produces poor-quality response.
9. Avoid fake premises. Do not begin a question with a premise with which respondents
may not agree, then ask about choices regarding it. Respondent who disagree with the
premise will be frustrated and not know how to answer. When did you stop beating
Your girl/boyfriend
10. Avoid asking about distant future intentions: After you graduate from college, get a
job, and are settled, will you invest a lot of money in the stock market
11. Avoid double negatives: Do you disagree with those who do not want to build a new
city swimming pool?
12. Avoid unbalanced response: Did you find the service at our hotel to be, Outstanding,
Excellent Superior or Good
Types of questionnaire items

More than one kind of questions such as for example multiple choice questions,
rating scales and open ended questions are likely to be found in a questionnaire.
In close ended questions, respondents are requested for their responses which

13
they may choose from ranges of possible responses. The other important aspect
of close ended questions (dichotomous, multiple choice and rating scale) is the
fact that they require less time to complete, and straightforward to code (for
computer analysis) and they do not discriminate unduly on the basis of how
articulate the respondents are.
However, these kind of questions do not enable respondents to express additional
opinions, add any remark, and explanations to the categories. Hence, the categories of
possible responses should be as exhaustive as possible. Now, let us briefly look at them.
Dichotomous questions: - Dichotomous questions require a ‘yes’/‘no’ response,
e.g. ‘have you ever had to appear in court?’, ‘do you prefer didactic methods to
child centered methods’? The dichotomous question is useful, because it entices
respondents to ‘come off the fence’ on an issue. Further, it is possible to code
responses quickly, there being only two categories of response. A dichotomous
question is also useful as a funneling or sorting device for subsequent questions,
for example: ‘if you answered “yes” to question X, please go to question Y; if
you answered “no” to question X, please go to question Z’. Scholars suggest that
if dichotomous questions are being used, then it is desirable to use several of
them to gain data on the same topic, in order to reduce the problems of
respondents’ ‘guessing’ answers.

On the other hand, the researcher must ask, for instance, whether a ‘yes’/‘no’ response
actually provides any useful information. Requiring respondents to make a ‘yes’/‘no’
decision may be inappropriate; it might be more appropriate to have a range of
responses, for example in a rating scale. There may be comparatively few complex or
subtle questions which can be answered with a simple ‘yes’ or ‘no’. A ‘yes’ or a ‘no’
may be inappropriate for a situation whose complexity is better served by a series of
questions which catch that complexity.

In addition to dichotomous questions (‘yes’/ ‘no’ questions) a piece of research

might ask for information about dichotomous variables, for example gender
(male/female), type of school (elementary/secondary), type of course
(vocational/non-vocational). In these cases only one of two responses can be
selected. This enables nominal data to be gathered, which can then be processed
using the chi-square statistic, the binomial test, the T-test, and cross-tabulations.
Multiple choice questions: Researchers can move towards multiple choice questions,
where the range of choices is designed to capture the likely range of responses to given
statements. For example, the researcher might ask a series of questions about a new
initiative on Service Delivery Improvement Program in Ethiopia. A set of responses to it
thus could be like the new initiative on Service Delivery Improvement Program in
Ethiopia is:
(a) A waste of time;
(b) An extra burden on the government;
(c) Not appropriate to our country;

14
(d) A useful complementary scheme;
(e) Well-planned and practicable.
The above categories should be discrete (i.e. having no overlap and being mutually
exclusive) and should exhaust the possible range of responses. Guidance needs to be
given on the completion of the multiple-choice, clarifying, for example, whether
respondents are able to tick only one response (a single answer mode)or several
responses (multiple answer mode)from the list. Like dichotomous questions, multiple
choice questions can be quickly coded and quickly aggregated to give frequencies of
response.

Rank ordering: - The rank order question is similar to the multiple-choice question in
that it identifies options from which respondents can choose. But rank ordering questions
moves beyond multiple choice items in that it asks respondents to identify priorities. This
enables a relative degree of preference, priority, intensity etc. to be charted. In the rank
ordering exercise a list of factors is set out and the respondent is required to place them in
a rank order. For example, respondents could be asked: Please indicate your priorities by
placing numbers in the boxes to indicate the ordering of your views,

1=the highest priority, 2=the second highest, and so on.

Whilst this might be enticing for the researcher, enabling fine distinctions possibly to be
made in priorities, it might be asking too much of the respondents to make such
distinctions. They genuinely might not be able to differentiate their responses, or they
simply might not feel strongly enough to make such distinctions. The inclusion of too
long a list might be overwhelming. Indeed Wilson and McLean (1994:26) suggest that it
is unrealistic to ask respondents to arrange priorities where there are more than five
ranks that have been requested. In the case of the list often points above, the researcher
might approach this problem in one of two ways. The list in the questionnaire item can
be reduced to five items only, in which case the range and comprehensiveness of
responses that fairly catches what the respondent feels is significantly reduced.

Alternatively, the list of ten items can be retained, but the request can be made to the
respondents only to rank their first five priorities, in which case the range is retained and
the task is not overwhelming (though the problem of sorting the data for analysis is
increased).Rankings are useful in indicating degrees of response. In this respect they are
like rating scales, discussed below.

Rating scales:- One way in which degrees of response, intensity of response, and the
move away from dichotomous questions has been managed can be seen in the notion of
rating scales— Likert scales, semantic differential scales, Thurston scales, Guttman
scaling. These are very useful devices for the researcher, as they build in a degree of
sensitivity and differentiation of response whilst still generating numbers.

A Likert scale (named after its deviser, Rensis Likert, 1932) provides a range of
responses to a given question or statement, for example: How important do you
consider work placements to be for secondary school students?

15
 1= not at all
2 = very little
3 = a little
4 = a lot
5 = a very great deal
All students should have access to free higher education.
1 = strongly disagree
2 = disagree
3 = neither agree nor disagree
4 = agree
5 = strongly agree
In these examples the categories need to be discrete and exhaust the range of possible
responses which respondents may wish to give. Notwithstanding the problems of
interpretation which arise as in the previous example—one respondent’s ‘agree’ may be
another’s ‘strongly agree’, one respondent’s ‘very little’ might be another’s ‘a little’—the
greater subtlety of response which is built into a rating scale renders this a very attractive
and widely used instrument in research. These two examples both indicate an important
feature of an attitude scaling instrument, viz. the assumption of unidimensionality in the
scale; the scale should only be measuring one thing at a time. Indeed this is a cornerstone
of Likert’s own thinking (1932).It is a very straightforward matter to convert a
dichotomous question into a multiple choice question. For example, instead of asking the
‘do you?’, ‘have you?’, ‘are you?’, ‘can you?’ type questions in a dichotomous format, a
simple addition to wording will convert it into a much more subtle rating scale, by
substituting the words ‘to what extent?’, ‘how far?’, ‘how much?’etc.
A semantic differential is a variation of a rating scale which operates by putting an
adjective at one end of a scale and its opposite at the other, for example:

How informative do you consider the new set of history text books to be?
Very fairly slightly neither slightly fairly very
___ ___ ___ ___ ___ ___
Useful Useless

The respondent indicates on the scale by circling or putting a mark on that position which
most represents what she or he feels.

The pioneers of this technique suggest that semantic differential scales are useful in three
contexts: evaluative (e.g. valuable-valueless, useful—useless, good—bad); potency (e.g.
large—small, weak— strong, light—heavy); and activity (e.g. quick—slow, active—
passive, dynamic-lethargic).
Rating scales are widely used in research, and rightly so, for they combine the opportunity
for a flexible response with the ability to determine frequencies, correlations and other
forms of quantitative analysis. They afford the researcher the freedom to fuse measurement
with opinion, quantity and quality.

Though rating scales are powerful and useful in research, the researcher, nevertheless,
needs to be aware of their limitations. For example, the researcher may not be able to infer

16
 a degree of sensitivity and subtlety from the data that they cannot bear. There are other
cautionary factors about rating scales, be they Likert scales or semantic differential scales:
There is no assumption of equal intervals between the categories, hence a rating of 4
indicates neither that it is twice as powerful as 2 nor that it is twice as strongly felt; one
cannot infer that the intensity of feeling in the Likert scale between ‘strongly disagree’ and
‘disagree’ somehow matches the intensity of feeling between ‘strongly agree’ and ‘agree’.
SOME EXAMPLES OF TYPES OF SURVEY QUESTIONS

1. Specific Information Request

In which year did you start the degree course? _____________

2. Category

Have you ever been or are you a student representative? (Tick which)

Yes (currently) 􀀀 Yes (in the past) 􀀀 Never 􀀀

3. Multiple Choice

Do you view the money you have spent on your higher education as any of the following? If so, tick which.

A luxury 􀀀 An investment 􀀀 A necessity 􀀀 A gamble 􀀀

A burden 􀀀 A right 􀀀 None of these

4. Scale

How would you describe your parent’s attitude to higher education at the time you applied? Please tick one of
the options below.

Very Positive Positive Mixed/Neutral Negative Very Negative Not Sure

􀀀 􀀀 􀀀 􀀀 􀀀 􀀀

5. Ranking

What do you see as the main purpose(s) of your degree study? Please rank all those relevant in order from 1
(most important) downwards:

Personal Development 􀀀 Career Advancement 􀀀

Subject Interest 􀀀 Recreation 􀀀

Fulfill Ambition 􀀀 Intellectual Stimulation 􀀀

Other 􀀀 (give details)……………………………………………

Open versus Closed Questions

There has long been a debate about open versus closed questions in survey research. An
open end (unstructured, free response) question asks a question (e.g., "what is your favorite
television program?’’) …to which respondents can give any answer. A closed-ended
(structured, fixed response) question both asks a question and gives the respondent fixed
responses from which to choose (e.g., "Is the president doing a very good, good, fair, or poor
job, in your opinion?").

17
Each form has advantages and disadvantages. The crucial issue is not which form is best.
Rather, it is under what conditions a form is most appropriate. A researcher’s choice to use
an open- or closed-ended question depends on the purpose and the practical limitations of a
research project. The demands of using open-ended questions, with interviewers writing
verbatim answers followed by time-consuming coding, may make them impractical for a
specific project.
Large-scale surveys have closed-ended questions because they are quicker and easier for both
respondents and researchers. Yet something important may be lost when an individual's
beliefs and feelings are forced into a few fixed categories that a researcher created. To learn
how a respondent thinks, to discover what is really important to him or her, or to get an
answer to a question with many possible answers (e.g. age), open questions may be best. In
addition, sensitive topics (e.g., sexual behavior, liquor consumption) may be more accurately
measured with closed questions.
The appropriateness of either close-ended or open-ended questions depends on a number of
factors. Researcher Paul Lazars Feld suggested that researchers use the following
considerations to determine appropriateness
1. The objective of the questionnaire: close-ended questions are suitable when the
researcher’s objective is to lead the respondents to express agreement or disagreement
with an explicit point of view, while open-ended questions are likely to be more
appropriate when the researcher wishes to learn how the respondent arrived at a particular
point of view.
2. The respondent’s level of information about the topic in question: open-ended questions
provide opportunities for the interviewer to ascertain alike of information on the part of
the respondent whereas close-ended questions do not.
3. The extent to which the topic has been thought through by the respondents: in situation
where the respondents have not yet crystallized their opinions, open-ended question is
preferable.
4. The ease with which respondents can communicate the content of the answer or the
extent to which respondents are motivated to communicate on the topic: close-ended
questions require less motivation to communicate on the part on the respondent.
The disadvantages of a question form can be reduced by mixing open-ended and closed-
ended questions in a questionnaire. Mixing them also offers a change of pace and helps
interviewers establish rapport. Periodic probes (i.e., follow-up questions by interviewers)
with closed-ended questions can reveal a respondent’s reasoning.

Advantage of Close- ended question Disadvantage of Close- ended question

 It is easier and quicker for  They can suggest ideas that the
respondents to answer. respondent would not otherwise
have
 The answers of different  Respondents with no opinion or
respondents are easier to no knowledge can answer
compare. anyway.

18
  Answers are easier to code and  Respondents can be frustrated
statistically analyze. because their desired answer is
not a choice.

 The response choices can clarify  It is confusing if many (e.g., 20)

question meaning for response choices are offered
respondents.

 Respondents are more likely to  Misinterpretation of a question

answer about sensitive topics. can go unnoticed

 There are fewer irrelevant or  Distinctions between respondent

confused answers to Questions answers may be blurred.

 Less articulate or less literate  Clerical mistakes or marking the

respondents are not at a wrong response is possible.
disadvantage.

 Replication is easier  They force respondents to give

simplistic responses to complex
issues.
 They force people to make
choices they would not make in
the real world

Advantages of Open Disadvantages of Open

▪ They permit an unlimited number ▪ Different respondents give
of possible answers. different degrees of detail in
answers.
▪ Respondents can answer in detail ▪ Responses may be irrelevant or
and can qualify and clarify buried in useless detail.
responses.
▪ Unanticipated findings can be ▪ Comparisons and statistical
discovered. analysis become very difficult.
▪ They permit adequate answers to ▪ Coding responses is difficult.
complex issues.
▪ They permit creativity, self- ▪ Articulate and highly literate
expression, and richness of detail. respondents have an advantage.
▪ They reveal a respondent's logic, ▪ Questions may be too general for
thinking process, and frame of respondent who lose direction:
reference

19
 ▪ Responses are written verbatim,
which is difficult for interviewers.
▪ A greater amount of respondent
time, thought, and effort is
necessary.
▪ Respondents can be intimidated
by questions.
▪ Answers take up a lot of space in
the questionnaire.
Developing Question Responses
There are three guiding principles when developing question responses.
Exhaustiveness (or inclusiveness)
Ensure that the response alternatives provide a sufficient range of responses to cover all
respondents. A question that asks about marital status and includes only ‘married’ and
‘divorced’ as alternatives is not providing an exhaustive set of options.
Numeric rating scales are a good way of providing an exhaustive set of responses for many
questions. Attitude questions should generally include a ‘don’t know’ or ‘no opinion’ option
so that those with no opinion are provided for. For some questions it is appropriate to add an
open category where respondents can create their own answer if the set provided has not been
exhaustive. For example, a question asking about the respondents’ country of birth might
provide a list of the most common countries but add a final, catch-all category of ‘Other
(please specify)’ to cover those respondents not covered by the preset responses.
Exclusiveness
This principle means that for each ‘question’ a person can provide one and only one answer
to the ‘question’. That is, the alternate responses should be mutually exclusive. This is not a
problem with rating scales and questions where the response alternatives are graded along a
single continuum. Respondents may have difficulty identifying where on the continuum they
lie but in principle they lie at a particular point. Exclusiveness is a problem where a person
might quite legitimately select more than one of the alternative responses. For example, in
your work place would you say that getting ahead is based on gender or merit? 1. Gender
2. Merit
Balancing categories
An imbalance of responses in one direction can produce bias (Payne, 1951). Where response
categories can be ordered from high to low there should be the same number of response
alternatives either side of what might be considered the neutral position. E.g. completely
approve, strongly approve, approve, neither approve nor disapprove, disapprove
The layout of the questionnaire

The format of a questionnaire is no less important than the wording of the questions asked.
An improperly formatted or laid out questionnaire could lead respondents to miss questions,
confuse respondents about the nature of the data desired and may lead respondents to throw
the questionnaire away. Moreover, respondents need to be introduced to the purposes of each

20
section of a questionnaire, so that they can become involved in it and maybe identify with it.
If space permits, it is useful to tell the respondent the purposes and foci of the sections/of the
questionnaire, and the reasons for the inclusion of the items.

Questionnaires as a general rule should be spread out and uncluttered. In other words, the
questionnaire should look easy, attractive and interesting rather than complicated, unclear,
forbidding and boring to respondents. Moreover, the questionnaire has to have an orderly
logical flow or sequence. For example, before asking respondents “how many children do
you have?”, it is better to ask him /her “Do you have children?” and then ask the number
of children the respondent has. A compressed layout is uninviting while a larger
questionnaire with sufficient space for questions and answers is more encouraging to
respondents. It is also advisable to use of high quality paper if funding permits.

Sectionalizing and sub lettering questions (e.g. Q9 (a) (b) (c)…) are useful techniques for
grouping together questions that deal with a specific issue. This practice is also a way of
making the questionnaire look smaller than it actually is which in turn encourages
respondents to fill out the questionnaire.

Moreover, it is important to include in the questionnaire, especially at the beginning,

assurances of confidentiality, anonymity, and non-traceability. This could be done indicating
that respondents need not give their name, that the data will be aggregated, that individuals
will not be able to be identified through the use of categories or details of their location etc.

Finally, a short note at the very end of the questionnaire can be added like: (a) asking
respondents to check that no answer has been inadvertently missed out; (b) soliciting an early
return of the completed schedule; (c) thanking respondents for their participation and co-
operation, and offer to send a short abstract of the major findings when the analysis is
completed.

However, inexperienced researchers fear that their questionnaire will look too long, and as a
result they squeeze several questions on to a single line abbreviate questions and try to use a
few pages as possible. But all these efforts are not advisable. When we put two Questions or
more on one line for example, we will cause some respondents to miss the second question
all together. Some respondents may misinterpret abbreviated questions.

In general respondents who find that they spent considerable time on the first page of what
seemed a short questionnaire will be more demoralized than respondents who quickly
complete the first several pages of what initially seems a rather long form. Moreover
respondents that look on spread out questions will make fewer errors and will not be forced to
reread confusing abbreviated questions.
Instructions
Another important element researchers must take into account when constructing a
questionnaire is instructions that go along with each question or with a set of questions.
Instructions should be included with any questions that are not self-explanatory: the

21
instructions must range from very simple ones such as “circle the appropriate category” to
more complex guidelines that explain how to set of priorities.
When an interviewer administers a questionnaire, the instructions are usually written for
him/her and thus are often short and concise, instructing the interviewer what to do when the
respondent provides a certain answer, when to probe for a more detail answer, or how to
clarify a certain question.
Covering letters/sheets and follow-up letters
A researcher prepares cover letters/sheets for the questionnaires to indicate the aim of the
research, to convey to respondents its importance, to assure them of confidentiality, and to
encourage their replies. Therefore, Choen L., Manion Lawrence and Morrison Keith, (2000 )
suggest that cover letter should:
• provide a title to the research;
• introduce the researcher, her/his name, address, organization, contact telephone/
fax/, e-mail address, together with an invitation to feel free to contact the researcher
for further clarification or details;
• indicate the purposes of the research; indicate the importance and benefits of the
research;
• indicate the address to which to return the questionnaire;
• indicate a return-by date;
• indicate any incentives for completing the questionnaire;
• Provide assurances of confidentiality, anonymity and non-traceability;
• Thank respondents in advance for their cooperation.
Piloting the questionnaire
Once you have constructed your questionnaire, you must pilot it. This means that you must
test it out to see if it is obtaining the result you require. This is done by asking people to read
it through and see if there are any ambiguities which you have not noticed. They should also
be asked to comment about the length, structure and wording of the questionnaire. Alter the
questions accordingly
Measures are seldom perfect and often indicate not only real differences but also artifact
difference, variations produced by the measuring procedure itself. Differences in
measurement scores that are due to anything other than real differences are termed as
measurement errors
The seriousness of the issues of validity and reliability are issues that prompted the
introduction of techniques for reducing measurement errors.
a. Validity: is concerned with the question “Am I measuring what I intent to measure?”
The concept of validity refers to quality and can be applied to any aspect of the research
process. With respect to measurement procedures it relates to whether a research instrument
is measuring what it set out to measure. There are two approaches used to establish the

22
validity of an instrument: the establishment of a logical link between the objectives of a study
and the questions used in an instrument, and the use of statistical analysis to demonstrate this
link.

b. Reliability: We use the word ‘reliability’ very often in our lives. When we say that a
person is reliable, what do we mean? We infer that s/he is dependable, consistent, predictable,
stable and honest.
The reliability of an instrument refers to its ability to produce consistent measurement each
time. When we administer an instrument under the same or similar conditions to the same
population and obtain similar results, we say that the instrument is ‘reliable’- the more similar
the results, the greater the reliability. You can look at reliability from two sides: reliability
(the extent of accuracy) and unreliability (the extent of inaccuracy). Ambiguity in the
wording of questions, a change in the physical setting for data collection, a respondent’s
mood when providing information, the nature of the interaction between interviewer and
interviewee and the regressive effect of an instrument are factors that can affect the reliability
of a research instrument.

Data Collection: Self-administered questionnaires and interviewing

Having formulated the research problem, developed a study design, constructed a research
instrument and selected a sample, you then collect the data from which you will draw
inferences and conclusions for your study. Depending upon your plans, you might commence
interviews, mail out a questionnaire, conduct experiments and/or make observations.
Types of questionnaire Administration
There are basically three ways of administering a questionnaire: the face-to-face interview,
the telephone interview and the self-administered questionnaire.
Face-to-face interviews
Using this method, trained interviewers administer the questionnaire personally to a
respondent. The interviewer records and codes answers as the interview proceeds. With
trained interviewers relatively complex questionnaires with complex branching patterns and
special modules can be administered. Since the interviewer is on the spot they can answer
respondent questions, clarify misunderstandings and probe answers to open-ended questions.
The presence of survey interviewer has several advantages compared to self-administered
questionnaire. First, interview survey ensures higher response rate than mail surveys. The
presence of an interviewer also decreases the number of “I don’t knows” answers. This is

23
done through probing by the interviewer- Probing in the technique used by the interviewer to
stimulate discussion and obtain more information.
Moreover, the presence of an interviewer creates the opportunity to clarify confusing
questions.
Guidelines for survey interview: - Though variation of specific details are likely, it possible
to outline some general guidelines for conducting survey interview. As a general rule, the
appearance and demeanor of the interviewer should not be offensive to the people being
interviewed. In other words, the interviewer should dress in a fashion similar to that of the
people he/she is interviewing. Familiarity with the questionnaire is the other point. This helps
the interviewer be able to read the questionnaire items without stumbling over words or
phrases. At times, respondent could give inappropriate answer. So, the interviewer is
responsible for probing the respondent.

Survey interview requires cautious coordination and control. It usually involves several
interviewers and these interviewers should themselves undergo training. Training session
should begin with the description of what the study is all about. It is also helpful to
accompany the questionnaire with specifications. Specifications are explanatory and
clarifying comments about handling difficult or confusing situations that may occur with
regard to specific questions in the questionnaire.

Telephone Interview
For long time, interviews were conducted in face to face situation. As telephones became
popular however, researchers began using telephone interviews. Telephone interview is
limited by definition to people who have telephones. But the necessity of having a
telephone line by respondents marginalizes poor people from being part of telephone
interview though this problem had been declining.
Telephone survey has many merits in that it saves the time and money of the researcher.
Moreover , respondents will be more honest in giving socially disapproved answers if
they don’t have to do it face to face. It may also be possible to probe them into more
sensitive areas.

The advantages of telephone interviews are briefly outlined. This technique:

• speeds up data collection; with the right resources, a large (over 1000 cases) nationwide
sample can be surveyed in as little as a few days
• Costs far less than face-to-face interviews: savings estimates range from 50% to75%
• enables computer recording of data to be carried out directly.
The disadvantages of the technique can be listed equally concisely:
- in the absence of personal contact, the interviewee feels less involved in the interview
and is therefore more likely to give superficial, random or hurried answers; moreover,
the interviewer will inevitably feel more detached (than in a face-to-face interview)
and may make less effort to ensure the success of the interview;

24
 - as the interviewer cannot see the respondent or his home, she cannot record
nonverbal data (such as the type of house, the interview environment, the social class
of the family as deduced from furnishings, etc.; in face-to-face interviews, such
features can be recorded);
- the limited amount of time available means that questions have to be simpler, more
concise and, in the final analysis, elementary
Of the various limitations listed here, the most serious are probably the first (lack of contact)
and the last (the pressure of time).The rapid pace of the interview allows little time for
thought; as already mentioned, short succinct questions prompt short succinct answers, as the
respondent perceives that he is required to answer briefly, immediately and without
hesitation. There are no pauses in the conversation; silence becomes embarrassing since the
respondent cannot use non-verbal signals to show that he is thinking about the answer.
Structures of Interview
Interviews can be grouped into two main types: 1. Structured 2. Semi-structured
a) Structured Interviews
Structured interviews involve the use of questionnaires based on a predetermined and
identical set of questions. The questions are usually read out by a researcher in a neutral tone
of voice to avoid influencing or prompting a particular response from a participant.
b) Semi-Structured Interviews
The interviewer will have a list of themes and areas to be covered and there may be some
standardized questions, but the interviewer may omit or add to some of these questions or
areas, depending on the situation and the flow of the conversation.
Self-administered questionnaires
The first advantage of the technique is obvious: an enormous saving on data-collection costs.
For example, in the case of a questionnaire handed out at the entrance to some public event (a
festival, exhibition, fair, etc.) and collected at the exit, a single field worker can gather
information from hundreds of people in a single day; to obtain similar results through face-to-
face interviews, dozens of interviewers would be kept busy for several days.

The main drawback to this technique is equally evident. In the techniques described earlier,
the questionnaire is filled in by a trained worker (the interviewer), who has had time to learn
the procedure and is unlikely to make gross mistakes. This is not so for the self-administered
questionnaire; many interviewees will be filling in a questionnaire for the first time, some
may never even have seen one before, and others may not know exactly what to do; the level
of education of the subjects will probably vary widely; moreover, many may not bother to fill
in the questionnaire, or to fill it in properly. As a result, self-administered questionnaires must
be as brief, concise and simple as possible.

25
Two basic cases of self-administration can be distinguished: group administration and
individual administration. An example of group administration may be seen when students in
a class are given questionnaires by an operator, who then gives instructions, assists with
compilation and collects the completed questionnaires at the end. In a case of this kind, the
problems mentioned are considerably reduced. The fact that the operator is on hand to
provide instructions and explanations makes gross error less likely; moreover, the operator
will ensure that all questionnaires are completed and handed in, thus avoiding the
phenomenon of self-selection. When group administration can be properly implemented, the
self-administered questionnaire is generally a valid technique.

Finally, to return to the question of individual completion of optional return questionnaires,

we will look in greater detail at the case of the mail questionnaire. This involves mailing
questionnaires to a sample of subjects who represent the whole population under
examination, together with a letter outlining the research and a stamped addressed envelope
for return.
The advantages and disadvantages of this technique are briefly listed:
Advantages:
• Very great savings on costs (the only immediate costs being those of printing and mailing);
• The questionnaire can be filled in at leisure, even at different times;
• Greater guarantee of anonymity than in a face-to-face interview (the returned questionnaire
does not contain any means of identification);
• No bias due to the presence of an interviewer;
• Subjects living far away or in isolated areas can be reached.
Disadvantages:
• The main problem with this technique is the low percentage of returns (often well below
50%), partly because there is no interviewer present to urge compliance;
• No control over completion of the questionnaire, which might be filled in by some other
member of the family, or by a secretary;

To conclude our discussion of telephone questionnaires and self-administered questionnaires,

it should be said that in both cases open questions are to be minimized. The time limits
imposed on telephone interviews oblige the respondent to answer quickly and briefly, leaving
no room for the inevitable hesitations and explanations that accompany open questions. In the

26
self-administered questionnaire, the absence of an interviewer removes one of the
prerequisites of open questions, which make sense only if there is an interviewer on hand to
guide, prompt and record. Without an interviewer, the open question lends itself to
misunderstanding and, especially, to omitted or incomplete transcription of responses.
Comparisons of the three data collection tools
As we have seen above, there are several ways to collect survey data. These ways have their
own strengths and weaknesses. Self-administered questionnaires are generally cheaper and
quicker than face to face interview. It also requires a small staff. In addition, respondents who
are reluctant to report controversial or deviant attitudes or behavior in interviews are willing
to respond to an anonymous self-administered questionnaire.

On the other hand, interviews also offer many advantages. Interview surveys produce few
incomplete questionnaires unlike as in the case of self-administered questionnaire. Further,
interview surveys are more effective for complicated issues though self-administered
questionnaire is more effective for sensitive issues.

When we look at telephone interview, it has benefit of saving time and money compared to
face to face interview. Telephone interviews are much cheaper and quicker. Moreover,
interviewer biases of face to face interview can be lessened when the interview is not seen by
the respondent. Therefore, researchers need to balance the advantages and disadvantages of
using anyone of the three survey research tools in line with their research needs and
resources.
Techniques for increasing response rate
There are techniques that can help improve response rates to postal or electronic
questionnaires:

 Follow-up calls (especially telephone reminders and special delivery letters).

 Pre-contact with respondents (telling them about the questionnaire).
 Type of postage (special delivery is superior to ordinary mail)
 Emphasizing Confidentiality: ensuring that all views to be published remain
anonymous, if appropriate
 Appeals to the respondent: based on the social, personal or other benefits that might
flow from the participation of a respondent

27
Chapter four
Processing and Analyzing Data
Processing and analyzing data involves a number of closely related operations which are
performed with the purpose of summarizing the collected data and organizing these in a
manner that they answer the research questions (objectives).

The analysis of quantitative data covers a wide range of things, from simple organization of
the data to complex statistical analysis. Quantitative information go through a process that is
primarily aimed at transforming the information into numerical values, called codes, so that
the information can easily be analyzed either manually or by computers.
The Data Processing operations are:
1. Editing- a process of examining the collected raw data to detect errors and omissions and
to correct these when possible.
2. Classification- a process of arranging data in groups or classes on the basis of common
characteristics. Depending on the nature of phenomenon involved
a) Classification according to attributes: here data is analyzed on the basis of common
characteristics which can either be
- Descriptive such as literacy, sex, religion etc. or
- Numerical such as weight, height, income etc.
Such classification can be either:
Simple classification: where we consider only one attribute, and divide the universe into two
classes—one class consisting of items possessing the given attribute and the other class
consisting of items which do not possess the given attribute.

Table 1. Hotel Employees with MBA Degree

Yes No Total
MBA Degree 21 9 30
Manifold classification: Here we consider two or more attributes simultaneously, and divide
the data into a number of classes.
Table 2. Educational Qualification of Hotel Employees
Yes No Total
M F M F M F
MBA Degree 12 9 3 6 15 15
B.Sc. H&HA 15 15 0 0 15 15
b) Classification according to class –intervals: is done with data relating to income, age,
weight, tariff, production, occupancy etc. Such quantitative data are known as the statistics of

28
variables and are classified on the basis of class –intervals. e.g. persons whose income are
within Rs 2001 to Rs 4000 can form one group or class, those with income within Rs 4001 to
Rs 6000 can form another group or class and so on.
The number of items which fall in a given class is known as the frequency of the given class.
Table 3. Pocket Money Received by IHM Students
Income Range Frequency %
Rs.1001-2000 10 50
Rs.2001-3000 8 40
Rs.3001-4000 2 10
Total 20 100
3. Tabulation-Tabulation is the process of summarizing raw data and displaying the same in
compact form for further analysis. It is an orderly arrangement of data in columns and rows.
Tabulation is essential because:
a) It conserves space and reduces explanatory and descriptive statement to a minimum.
b) It facilitates the process of comparison.
c) It facilitates the summation of items and the detection of errors and omissions.
d) It provides the basis for various statistical computations.

Tabulation may also be classified as simple and complex tabulation. Simple tabulation
generally results in one-way tables which supply answers to questions about one
characteristic of data only. Complex tabulation usually results on two-way tables (which give
information about two inter-related characteristics of data), three –way tables or still higher
order tables, also known as manifold tables.

29