1 DATA

1 Data

1.1 Data Collection Methods

The first step in any investigation (inquiry) is data collection. Information can either be collected directly or

indirectly from the entire population or a sample. There are many methods of collecting data which includes the

ones illustrated in the flow chart below

Experimental methods are called so because in them the investigator in a laboratory tests the hypothesis about

the cause and effect relationship by manipulating the independent variables under controlled conditions. Non-

Experimental methods are called so because in them the investigator does not control or change any aspect of

1
1.1 Data Collection Methods 1 DATA

the situation under study but simply describes what naturally occurs at a certain point or period of time. Non-

Experimental methods are widely used in social sciences. Some of the Non-Experimental methods used for data

collection are outlined below.

1. Field study:- aims at testing hypothesis in natural life situations. It differs from field experiment in that the

researcher does not control or manipulate the independent variables but both of them are carried out in natural

conditions.

Merits:

(a) The method is realistic as it is carried out in natural conditions

(b) It’s easy to obtain data with large number of variables.

Demerits

(a) Independent variables are not manipulated.

(b) Co-operation of the organization is often difficult to obtain.

(c) Data is likely to contain unknown sampling biasness.

2
1.1 Data Collection Methods 1 DATA

(d) The dross rate (proportion of irrelevant data) may be high in such studies.

(e) Measurement is not precise as in laboratory because of influence of confounding variables.

2. Census. A census is a study that obtains data from every member of a population (totality of individuals

/items pertaining to certain characteristics). In most studies, a census is not practical, because of the cost

and/or time required.

3. Sample survey. A sample survey is a study that obtains data from a subset of a population, in order to

estimate population attributes/ characteristics. Surveys of human populations and institutions are common in

government, health, social science and marketing research.

4. Case study –It’s a method of intensively exploring and analyzing the life of a single social unit be it a family,

person, an institution, cultural group or even an entire community. In this method no attempt is made to

exercise experimental or statistical control and phenomena related to the unit are studied in natural. The

researcher has several discretion in gathering information from a variety of sources such as diaries, letters,

autobiographies, records in office, files or personal interviews.

Merits:

3
1.1 Data Collection Methods 1 DATA

(a) The method is less expensive than other methods.

(b) Very intensive in nature –aims at studying a few units rather than several

(c) Data collection is flexible since the researcher is free to approach the problem from any angle.

(d) Data is collected from natural settings.

Demerits

(a) It lacks internal validity which is basic to scientific evidence.

(b) Only one unit of the defined population is studied. Hence the findings of case study cannot be used as

abase for generalization about a large population. They lack external validity.

(c) Case studies are more time consuming than other methods.

5. Experiment. An experiment is a controlled study in which the researcher attempts to understand cause-

and-effect relationships. In experiments actual experiment is carried out on certain individuals / units about

whom information is drawn. The study is "controlled" in the sense that the researcher controls how subjects

are assigned to groups and which treatments each group receives.

4
1.2 Population and Sample Population: 1 DATA

6. Observational study. Like experiments, observational studies attempt to understand cause-and-effect re-

lationships. However, unlike experiments, the researcher is not able to control how subjects are assigned to

groups and/or which treatments each group receives. Under this method, information is sought by direct

observation by the investigator.

1.2 Population and Sample Population:

Population refers to the entire set of individuals about which findings of a survey refer to while a Sample refers to

a subset of population selected for a study. Sample Design refers to the scheme by which items are chosen for the

sample. Sample unit refers to the element of the sample selected from the population. Unit of analysis refers to a

unit at which analysis will be done for inferring about the population.

Sampling Frames

For probability sampling, we must have a list of all the individuals (units) in the population. This list or sampling

frame is the basis for the selection process of the sample. A sampling frame is a clear and concise description of the

population under study, by virtue of which the population units can be identified unambiguously and contacted, if

desired, for the purpose of the survey. Based on the sampling frame, the sampling design could also be classified

5
1.2 Population and Sample Population: 1 DATA

as:

Individual Surveys; if List of individuals is available or when the size of population is small Special population.

Household Surveys; If it’s Based on the census of the households and if the individual level information is

unlikely to be available In practice, it’s limited to small geographical areas and known as “area sampling frame”

e.g. Demographic and Health Surveys (DHS).

Institutional Surveys; If it’s Based on the census of say Hospital/clinic lists eg

• 1990 National Hospital Discharge Survey

• National Ambulatory Medical Care Survey

Problems of Sampling Frame

1. Missing elements

2. Noncoverage

3. Incomplete frame

4. Old list

6
1.3 Sampling 1 DATA

5. Undercoverage

6. May not be readily available

7. Expensive to gather

1.3 Sampling

Sampling is a statistical process of selecting a representative sample. We have probability sampling and non-

probability sampling. Probability Sampling involve a mathematical chance of selecting the respondent. Every

unit in the population has a chance, greater than zero, of being selected in the sample. Thus producing unbiased

estimates. They include;

1. Simple random sampling

2. Systematic sampling

3. Stratified sampling

4. Cluster sampling andmulti-stage sampling

7
1.3 Sampling 1 DATA

Non-probability sampling on the other hand is any sampling method where some elements of the population have

no chance of selection (also referred to as “out of coverage”/”undercovered”), or where the probability of selection

can’t be accurately determined. It yields a non-random sample therefore making it difficult to extrapolate from

the sample to the population. They include; Judgemental sampling, purposive sampling, convenience sampling

(subjective) and Snow-ball sampling (rare group/disease study)

8
1.3 Sampling 1 DATA

1.3.1 Sampling Procedure

Sampling involves two tasks; How to select the elements and how to estimate the population characteristics from

the sampling units. We employ some randomization process for sample selection so that there is no preferential

treatment in selection which may introduce selectivity bias.

1.3.2 Reasons Behind sampling

1. Cost; the sample can furnish data of sufficient accuracy at much lower cost.

2. Time; the sample provides information faster than census thus ensuring timely decision making.

3. Accuracy; it is easier to control data collection errors in a sample survey as opposed to census.

4. Risky or destructive test call for sample survey not census eg testing a new drug.

1.3.3 Probability Sampling Techniques

A. Simple Random Sampling (SRS).

In this design, each element has an equal probability of being selected from a list of all population units (sample of

n from N population). Though it’s attractive for its simplicity, the design is not usually used in the sample survey

in practice for several reasons:

9
1.3 Sampling 1 DATA

1. Lack of listing frame: the method requires that a list of population elements be available, which is not the

case for many populations.

2. Problem of small area estimation or domain analysis: For a small sample from a large population, all the

areas may not have enough sample size for making small area estimation or for domain analysis by variables of

interest.

3. Not cost effective: SRS requires covering of whole population which may reside in a large geographic area;

interviewing few samples spread sparsely over a large area would be very costly.

Implementation of SRS sampling:

(a) Listing (sampling) Frame.

(b) Random number table (from published table or computer generated)

(c) Selection of sample

10
1.3 Sampling 1 DATA

B. Systematic Sampling:

Systematic sampling, either by itself or in combination with some other method, may be the most widely used

method of sampling. In systematic sampling we select samples “evenly” from the list (sampling frame): First, we

consider that we are dividing the list evenly into some “blocks”. Then, we select a sample element from each

block. In systematic sampling, only the first unit is selected at random, the rest being selected according to a

predetermined pattern. To select a systematic sample of n units, the first unit is selected with a random start r
N
from 1 to k sample, where k = n sample intervals, and after the selection of first sample, every k th unit is included

where 1 ≤ r ≤ k.

11
1.3 Sampling 1 DATA

100
For example: Let N = 100, n = 10, then k = 10 . Then the random start r is selected between 1 and 10 (say,

r = 7). So, the sample will be selected from the population with serial indexes of: 7, 17, 27, ......, 97.i.e., r, r +

k, r + 2k, ......., r + (n − 1)k.

N
What could be done if k = n is not an integer?
1000
Consider, n = 175 and N = 1000. So, k = 175 = 5.71. One of the solutions is to make k rounded off to an
1000 1000
integer, i.e.,k = 5 or k = 6. Now, if k = 5, then n = 5 = 200; or, If k = 6, then n = 6 = 166.67 167.

Which n should be chosen?

if k = 5 is considered, stop the selection of samples when n = 175 achieved and if k = 6 is considered, treat

the sampling frame as a circular list and continue the selection of samples from the beginning of the list after

exhausting the list during the first cycle.

An alternative procedure is to keep k non-integer and continue the sample selection as follows: Let us consider,k =

5.71, and r = 4. So, the first sample is 4th in the list. The second = (4 + 5.71) = 9.71 9th in the list, the third

= (4 + 2 × 5.71) = 15.42 15th in the list, and so on. (The last sample is: 4 + 5.71(175 − 1) = 997.54 997th in

the list). Note that, k is switching between 5 and 6.

12
1.3 Sampling 1 DATA

Advantages:

Systematic sampling has many merits:

1. Provides a better random distribution than SRS

2. Simple to implement

3. May be started without a complete listing frame (say, interview of every 9th patient coming to a clinic).

4. With ordered list, the variance may be smaller than SRS (see below for exceptions

Disadvantages:

1. Periodicity (cyclic variation)

2. linear trend

When to use systematic sampling?

1. Even preferred over SRS.

2. When no list of population exists.

13
1.3 Sampling 1 DATA

3. When the list is roughly of random order.

4. Small area/population.

C. Stratified Sampling

In stratified sampling the population is partitioned into groups, called strata, and sampling is performed separately

within each stratum. This sampling technique is used when;

1. Population groups may have different values for the responses of interest.

2. we want to improve our estimation for each group separately.

3. To ensure adequate sample size for each group.

In stratified sampling designs:

1. Stratum variables are mutually exclusive (no over lapping), e.g., urban/rural areas, economic categories, geo-

graphic regions, race, sex, etc. The principal objective of stratification is to reduce sampling errors.

2. The population (elements) should be homogenous within-stratum, and the population (elements) should be

heterogeneous between the strata.

14
1.3 Sampling 1 DATA

Advantages

1. Provides opportunity to study the stratum; variations - estimation could be made for each stratum

2. Disproportionate sample may be selected from each stratum

3. The precision is likely to increase as variance may be smaller than simple random case with same sample size

4. Field works can be organized using the strata (e.g., by geographical areas or regions)

5. Reduce survey costs.

Disadvantages

1. Sampling frame is needed for each stratum

2. Analysis method is complex

3. Correct variance estimation

4. Data analysis should take sampling “weight” into account for disproportionate sampling of strata

5. Sample size estimation is difficult in practice.

15
1.3 Sampling 1 DATA

Allocation of Stratified Sampling

The major task of stratified sampling design is the appropriate allocation of samples to different strata. Types of

allocation methods:

• Equal allocation

• Proportional to stratum size

• Cost based sample allocation

Equal Allocation

Divide the number of sample units n equally among the K strata. i.e. ni = nk . For example, given n = 100 and
100
k = 4 strata then ni = 4 = 25 units in each stratum.

Disadvantages of equal allocation:

May need to use weighting to have unbiased estimates

Proportional allocation

Make the proportion of each stratum sampled identical to the proportion of the population that is let the sample

16
1.3 Sampling 1 DATA

fraction be f = n
N. So, ni = f Ni = n NNi , Where Ni
N is the stratum weight.
100
For example, givenN = 1000, n = 100 then f = 1000 = 0.1 now suppose N1 = 700 and N2 = 300 then

n1 = 700 × 0.1 = 70 and n2 = 300 × 0.1 = 30.

Disadvantage of proportional allocation:

Sample size in a stratum may be low thus providing unreliable stratum-specific results.

D. Cluster Sampling

In many practical situations the population elements are grouped into a number of clusters. A list of clusters can

be constructed as the sampling frame but a complete list of elements is often unavailable, or too expensive to

construct. In this case it is necessary to use cluster sampling where a random sample of clusters is taken and some

or all elements in the selected clusters are observed. Cluster sampling is also preferable in terms of cost, because it

is much cheaper, easier and quicker to collect data from adjoining elements than elements chosen at random. On

the other hand, cluster sampling is less informative and less efficient per elements in the sample, due to similarities

of elements within the same cluster. The loss of efficiency, however, can often be compensated by increasing the

overall sample size. Thus, in terms of unit cost, the cluster sampling plan is efficient.

17
1.3 Sampling 1 DATA

E. Multi-Stage Sampling

Here the respondents are chosen through a process of defined stages. For instance, residents within Kibera (Nairobi)

may have been chosen for a survey through the following process: Throughout the country (Kenya), the Nairobi

City may have been selected at random, ( stage 1), within Nairobi, Langata (constituency) is selected again at

random (stage 2), Kibera is then selected within Langata (stage 3), then polling stations from Kibera (stage 4)

and then individuals from the electoral voters’ register (stage 5). As demonstrated in the five stages gone through

before the final selection of respondents were selected from the electoral voters’ register.

Advantages of probability sampling

1. Provides a quantitative measure of the extent of variation due to random effects

2. Provides data of known quality

3. Provides data in timely fashion

4. Provides acceptable data at minimum cost

5. Better control over nonsampling sources of errors

18
1.3 Sampling 1 DATA

6. Mathematical statistics and probability can be applied to analyze and interpret the data

1.3.4 Non-probability Sampling

Social research is often conducted in situations where a researcher cannot select the kinds of probability samples

used in large-scale social surveys. For example, say you wanted to study homelessness - there is no list of homeless

individuals nor are you likely to create such a list. However, you need to get some kind of a sample of respondents

in order to conduct your research. To gather such a sample, you would likely use some form of non-probability

sampling. There are four primary types of non-probability sampling methods:

A. Convinience Sampling

It is a method of choosing subjects who are available or easy to find. This method is also sometimes referred to

as haphazard, accidental, or availability sampling. The primary advantage of the method is that it is very easy to

carry out, relative to other methods.

Advantages of convenience sampling

1. It is very easy to carry out with few rules governing how the sample should be collected.

2. The relative cost and time required to carry out a convenience sample are small in comparison to probability

19
1.3 Sampling 1 DATA

sampling techniques. This enables you to achieve the sample size you want in a relatively fast and inexpensive

way.

3. The convenience sample may help you gather useful data and information that would have been impossible

using probability sampling techniques, which require more formal access to lists of populations. For example,

imagine you were interested in understanding more about employee satisfaction in a single, large organisation

in the United States. You intended to collect your data using a questionnaire. The manager who has kindly

given you access to conduct your research is unable to get permission to get a list of all employees in the

organisation, which you would need to use a probability sampling technique such as simple random sampling

or systematic random sampling. However, the manager has managed to secure permission for you to spend

two days in the organisation to collect as many questionnaire responses as possible. You decide to spend

the two days at the entrance of the organisation where all employees have to pass through to get to their

desks. Whilst a probability sampling technique would have been preferred, the convenience sample was the

only sampling technique that you could use to collect data. Irrespective of the disadvantages of convenience

sampling, discussed below, without the use of this sampling technique, you may not have been able to get

access to any data on employee satisfaction in the organisation.

20
1.3 Sampling 1 DATA

Demerits

1. One can never be certain what population the participants in the study represent. The population is unknown.

2. The method is haphazard, and the cases studied probably don’t represent any population you could come up

with. However, its very useful for pilot studies.

3. The convenience sample often suffers from a number of biases. This is because a convenience sample can

lead to the under-representation or over-representation of particular groups within the sample. If we take

the large organisation: It may be that the organisation has multiple sites, with employee satisfaction varying

considerably between these sites. By conducting the survey at the headquarters of the organisation, we may

miss the differences in employee satisfaction amongst those at different sites, including non-office workers. We

also do not know why some employees agreed to take part in the survey, whilst others did not. Was it because

some employees were simply too busy? Did they not trust the intentions of the survey? Did others take part

out of kindness or because they had a particular grievance with the organisation? These types of biases are

quite typical in convenience sampling. Since the sampling frame is not known, and the sample is not chosen

at random, the inherent bias in convenience sampling means that the sample is unlikely to be representative

21
1.3 Sampling 1 DATA

of the population being studied. This undermines your ability to make generalisations from your sample to the

population you are studying.

B. Quota Sampling

Quota sampling is designed to overcome the most obvious flaw of availability sampling. Rather than taking just

anyone, you set quotas to ensure that the sample you get represents certain characteristics in proportion to their

prevalence in the population. Note that for this method, you have to know something about the characteristics of

the population ahead of time. Say you want to make sure you have a sample proportional to the population in terms

of gender - you have to know what percentage of the population is male and female, then collect sample until yours

matches. Marketing studies are particularly fond of this form of research design. The primary problem with this

form of sampling is that even when we know that a quota sample is representative of the particular characteristics

for which quotas have been set, we have no way of knowing if sample is representative in terms of any other

characteristics. If we set quotas for gender and age, we are likely to attain a sample with good representativeness

on age and gender, but one that may not be very representative in terms of income and education or other factors.

Moreover, because researchers can set quotas for only a small fraction of the characteristics relevant to a study

quota sampling is really not much better than availability sampling. To reiterate, you must know the characteristics

22
1.3 Sampling 1 DATA

of the entire population to set quotas; otherwise there’s not much point to setting up quotas. Finally, interviewers

often introduce bias when allowed to self-select respondents, which is usually the case in this form of research. In

choosing males 18-25, interviewers are more likely to choose those that are better-dressed, seem more approachable

or less threatening. That may be understandable from a practical point of view, but it introduces bias into research

findings. Imagine that a researcher wants to understand more about the career goals of students at a single

university. Lets say that the university has roughly 10,000 students. suppose we were interested in comparing the

differences in career goals between male and female students at the single university. If this was the case, we would

want to ensure that the sample we selected had a proportional number of male and female students relative to the

population. To create a quota sample, there are three steps:

• Choose the relevant grouping criteria and divide the population accordingly.

• Calculate a quota (number of units that should be included in each group)

• Continue to invite units until the quota for each group is met.

Advantages of quota sampling

1. It particularly useful when you are unable to obtain a probability sample, but you are still trying to create a

23
1.3 Sampling 1 DATA

sample that is as representative as possible of the population being studied. In this respect, it is the non-

probability based equivalent of the stratified random sample.

2. Unlike probability sampling techniques, especially stratified random sampling, quota sampling is much quicker

and easier to carry out because it does not require a sampling frame and the strict use of random sampling

techniques.

3. The quota sample improves the representation of particular strata (groups) within the population, as well as

ensuring that these strata are not over-represented. For example, it would ensure that we have sufficient male

students taking part in the research (60% of our sample size of 100; hence, 60 male students). It would also

make sure we did not have more than 60 male students, which would result in an over-representation of male

students in our research.

4. It allows comparison of groups.

Disadvantages of quota sampling

1. In quota sampling, the sample has not been chosen using random selection, which makes it impossible to

determine the possible sampling error.

24
1.3 Sampling 1 DATA

2. there is sampling bias. Thus nostatistical inferences from the sample to the population. This can lead to

problems of external validity.

3. Also, with quota sampling it must be possible to clearly divide the population into strata; that is, each unit

from the population must only belong to one stratum. In our example, this would be fairly simple, since our

strata are male and female students. Clearly, a student could only be classified as either male or female. No

student could fit into both categories (ignoring transgender issues).

Purposive Sampling

Purposive sampling is a sampling method in which elements are chosen based on purpose of the study. Purposive

sampling may involve studying the entire population of some limited group or a subset of a population. As with

other non-probability sampling methods, purposive sampling does not produce a sample that is representative of a

larger population, but it can be exactly what is needed in some cases - study of organization, community, or some

other clearly defined and relatively limited group.

Advantages of purposive sampling

1. There are a wide range of qualitative research designs that researchers can draw on. Achieving the goals of

25
1.3 Sampling 1 DATA

such qualitative research designs requires different types of sampling strategy and sampling technique. One of

the major benefits of purposive sampling is the wide range of sampling techniques that can be used across such

qualitative research designs; purposive sampling techniques that range from homogeneous sampling through

to critical case sampling, expert sampling, and more.

2. Whilst the various purposive sampling techniques each have different goals, they can provide researchers with

the justification to make generalisations from the sample that is being studied, whether such generalisations

are theoretical, analytic and/or logical in nature.

3. Qualitative research designs can involve multiple phases, with each phase building on the previous one. In such

instances, different types of sampling technique may be required at each phase. Purposive sampling is useful

in these instances because it provides a wide range of non-probability sampling techniques for the researcher

to draw on. For example, critical case sampling may be used to investigate whether a phenomenon is worth

investigating further, before adopting an expert sampling approach to examine specific issues further.

Disadvantages of purposive sampling

1. Purposive samples, irrespective of the type of purposive sampling used, can be highly prone to researcher

26
1.3 Sampling 1 DATA

bias. The idea that a purposive sample has been created based on the judgement of the researcher is not a

good defence when it comes to alleviating possible researcher biases, especially when compared with probability

sampling techniques that are designed to reduce such biases. However, this judgemental, subjective component

of purpose sampling is only a major disadvantage when such judgements are ill-conceived or poorly considered;

that is, where judgements have not been based on clear criteria, whether a theoretical framework, expert

elicitation, or some other accepted criteria.

2. The subjectivity and non-probability based nature of unit selection (i.e. selecting people, cases/organisations,

etc.) in purposive sampling means that it can be difficult to defend the representativeness of the sample. In

other words, it can be difficult to convince the reader that the judgement you used to select units to study

was appropriate. For this reason, it can also be difficult to convince the reader that research using purposive

sampling achieved theoretical/analytic/logical generalisation. After all, if different units had been selected,

would the results and any generalisations have been the same.

D. Snowball Sampling

Snowball sampling is a method in which a researcher identifies one member of some population of interest, speaks

27
1.3 Sampling 1 DATA

to him/her, and then asks that person to identify others in the population that the researcher might speak to. This

person is then asked to refer the researcher to yet another person, and so on. Snowball sampling is very good for

cases where members of a special population are difficult to locate. For example, populations that are subject to

social stigma and marginalisation, such as suffers of AIDS/HIV, as well as individuals engaged in illicit or illegal

activities, including prostitution and drug use. Snowball sampling is useful in such scenarios because: the method

creates a sample with questionable representativeness. A researcher is not sure who is in the sample. In effect

snowball sampling often leads the researcher into a realm he/she knows little about. It can be difficult to determine

how a sample compares to a larger population. Also, there’s an issue of who respondents refer you to - friends refer

to friends, less likely to refer to ones they don’t like, fear, etc. Snowball sampling is a useful choice of sampling

strategy when the population you are interested in studying is hidden or hard-to-reach.

Advantages of Snowball Sampling

1. The chain referral process allows the researcher to reach populations that are difficult to sample when using

other sampling methods.

2. The process is cheap, simple and cost-efficient.

28
1.3 Sampling 1 DATA

3. This sampling technique needs little planning and fewer workforce compared to other sampling techniques.

Disadvantages of Snowball Sampling

1. The researcher has little control over the sampling method. The subjects that the researcher can obtain rely

mainly on the previous subjects that were observed.

2. Representativeness of the sample is not guaranteed. The researcher has no idea of the true distribution of the

population and of the sample.

3. Sampling bias is also a fear of researchers when using this sampling technique. Initial subjects tend to nominate

people that they know well. Because of this, it is highly possible that the subjects share the same traits and

characteristics, thus, it is possible that the sample that the researcher will obtain is only a small subgroup of

the entire population.

1.3.5 Limitations of Sampling

1. Sampling frame: may need complete enumeration.

2. Errors of sampling may be high in small areas.

29
1.3 Sampling 1 DATA

3. May not be appropriate for the study objectives/questions

4. Representativeness may be vague, controversial.

1.3.6 Characteristics of Good sampling

A good sample should;

• Meet the requirements of the study objectives.

• Provides reliable results.

• Clearly understandable.

• Manageable/realistic: could be implemented.

• Time consideration: reasonable and timely.

• Cost consideration: economical.

• Interpretation: accurate, representative.

• Acceptability

30
1.4 Survey Administration 1 DATA

1.4 Survey Administration

1.4.1 Steps in Survey

1. Setting the study objectives; What are the objectives of the study? Is survey the best procedure to collect

data? Why other study design (experimental, quasi-experimental, community randomized trials, epidemiologic

designs„e.g., case-control study) is not appropriate for the study? What information/data need to be collected?

2. Defining the study population; Representativeness Sampling frame.

3. Decide sample design: alternative considerations.

4. Questionnaire design; Appropriateness, acceptability, culturally appropriate, understandable Pre-test: Appro-

priate, acceptable, culturally appropriate, will answer.

5. Fieldwork; Training/Supervision Quality monitoring Timing: seasonality.

6. Quality assurance in every steps minimizing errors/bias/cheating.

7. Data entry/compilation Validation Feedback.

8. Analysis: Design consideration.

31
1.4 Survey Administration 1 DATA

9. Dissemination.

10. Plans for next survey: what did you learn, what did you miss?

1.4.2 Modes of Survey Administration

1. Self-Administered Surveys

2. Personal interview

3. Telephone

4. Mail

5. Computer assisted self-interviewing(CASI) Variants: CAPI (personal interview); CATI (telephone interview) –

Replaces the papers.

6. Combination of methods.

1. Self-Administered Surveys

Self-administered surveys have special strengths and weaknesses. They are useful in describing the characteristics

of a large population and make large samples feasible.

32
1.4 Survey Administration 1 DATA

Advantages:

1. Low cost. Extensive training is not required to administer the survey. Processing and analysis are usually

simpler and cheaper than for other methods.

2. Reduction in biasing error. The questionnaire reduces the bias that might result from personal characteristics

of interviewers and/or their interviewing skills.

3. Greater anonymity. Absence of an interviewer provides greater anonymity for the respondent. This is especially

helpful when the survey deals with sensitive issues.

4. Convenience to the respondents (may complete any time at his/her own convenient time).

5. Accessibility (greater coverage, even in the remote areas)

6. May provide more reliable information (e.g. may consult with others or check records to avoid recall bias)

Disadvantages:

1. Requires simple questions. The questions must be straightforward enough to be comprehended solely on the

basis of printed instructions and definitions.

33
1.4 Survey Administration 1 DATA

2. No opportunity for probing. The answers must be accepted as final. Researchers have no opportunity to clarify

ambiguous answers.

3. Low response rate; respondents may not respond to all questions and/or may not return questionnaire.

4. The respondent must be literate to read and understand the questionnaire.

5. Introduce self selection bias.

6. Not suitable for complex questionnaire.

2. Interview Surveys

Unlike questionnaires interviewers ask questions orally and record respondents’ answers. This type of survey generally

decreases the number of ; ”do not know” and ”no answer” responses, compared with self-administered surveys.

Interviewers also provide a guard against confusing items. If a respondent has misunderstood a question, the

interviewer can clarify, thereby obtaining relevant responses.

Interviewer selection: background characteristics (race, sex, education, culture) listening skill recording skill

experience unbiased observation/recording.

Interviewer training: be familiar with the study objectives and significance thorough familiarity with the

34
1.4 Survey Administration 1 DATA

questionnaire contextual and cultural issues privacy and confidentiality informed consent and ethical issues unbiased

view mock interview session.

Supervision of the interviewer: Spot check Questionnaire check Reinterview (reliability check)

Advantages

1. Flexibility: Allows flexibility in the questioning process and allows the interviewer to clarify terms that are

unclear.

2. Control of the interview situation: Can ensure that the interview is conducted in private, and respondents do

not have the opportunity to consult one another before giving their answers.

3. High response rate: Respondents who would not normally respond to a mail questionnaire will often respond

to a request for a personal interview.

4. May record non-verbal behaviour, activities, facilities, contexts.

5. Complex questionnaire may be used.

6. Illiterate respondents may participate.

35
1.4 Survey Administration 1 DATA

Disadvantages

1. Higher cost. Costs are involved in selecting, training, and supervising interviewers; perhaps in paying them;

and in the travel and time required to conduct interviews.

2. Interviewer bias. The advantage of flexibility leaves room for the interviewer’s personal influence and bias,

making an interview subject to interviewer bias.

3. Lack of anonymity. Often the interviewer knows all or many of the respondents. Respondents may feel

threatened or intimidated by the interviewer, especially if a respondent is sensitive to the topic or to some of

the questions.

4. Less accessibility.

5. Inconvenience.

6. Often no opportunity to consult records, families, relatives.

36
1.4 Survey Administration 1 DATA

C. Telephone Interview
Advantages:

1. Less expensive.

2. Shorter data collection period than personal interviews.

3. Better response than mail surveys.

Disadavantages

1. Biased against households without telephone, unlisted number.

2. Nonresponse

3. Difficult for sensitive issues or complex topics.

4. Limited to verbal responses.

E. Focus Groups

Focus groups are useful in obtaining a particular kind of information that would be difficult to obtain using other

methodologies. A focus group typically can be defined as a group of people who possess certain characteristics and

37
1.4 Survey Administration 1 DATA

provide information of a qualitative nature in a focused discussion. Focus groups generally are composed of six

to twelve people. Size is conditioned by two factors: the group must be small enough for everyone to participate,

yet large enough to provide diversity. This group is special in terms of purpose, size, composition, and procedures.

Participants are selected because they have certain characteristics in common that relate to the topic at hand, such

as parents of gang members, and, generally, the participants are unfamiliar with each other. Typically, more than

one focus group should be convened, since a group of seven to twelve people could be too atypical to offer any

general insights on the gang problem. A trained moderator probes for different perceptions and points of view,

without pressure to reach consensus. Focus groups have been found helpful in assessing needs, developing plans,

testing new ideas, or improving existing programs.

Advantages:

1. Flexibility; allows the moderator to probe for more in-depth analysis and ask participants to elaborate on their

responses.

2. Outcomes are quickly known.

3. They may cost less in terms of planning and conducting than large surveys and personal interviews.

38
1.4 Survey Administration 1 DATA

Limitations

1. A skilled moderator is essential.

2. Differences between groups can be troublesome to analyze because of the qualitative nature of the data.

3. Groups are difficult to assemble. People must take the time to come to a designated place at a particular time.

4. Participants may be less candid in their responses in front of peers.

New Terminology in Computer Age

PPI: Paper and Pencil Interview

CAI: Computer-Assisted Interview

CATI: Computer-Assisted Telephone Interview

CAPI: Computer-Assisted Personal Interview

CASI: Computer-Assisted Self-Interview

Internet Surveys: Surveys over the www.

39
1.5 Sample Size Determination 1 DATA

1.5 Sample Size Determination

Sample Size Determination is influenced by factors like the purpose of the study, population size, the risk of selecting

a "bad" sample, and the allowable sampling error. There are several approaches to determining the sample size.

These include using a census for small populations, imitating a sample size of similar studies, using published tables,

and applying formulas to calculate a sample size.

Using a census for small populations

One approach is to use the entire population as the sample. It’s impractical for large populations. A census

eliminates sampling error and provides data on all the individuals in the population. Finally, virtually the entire

population would have to be sampled in small populations to achieve a desirable level of precision

Using a sample size of a similar study

Another approach is to use the same sample size as those of studies similar to the one you plan. Without reviewing

the procedures employed in these studies you may run the risk of repeating errors that were made in determining

the sample size for another study. However, a review of the literature in your discipline can provide guidance about

"typical" sample sizes which are used.

40
1.5 Sample Size Determination 1 DATA

Using published tables

One can also rely on published tables which provide the sample size for a given set of criteria. Yamane (1967)

Table 2.1 and Table 2.2 present sample sizes that would be necessary for given combinations of precision, confidence

levels, and variability.

Note:

• these sample sizes reflect the number of obtained responses, and not necessarily the number of surveys mailed

or interviews planned (this number is often increased to compensate for non-response).

• the sample sizes in Table 2.2 presume that the attributes being measured are distributed normally or nearly so.

If this assumption cannot be met, then the entire population may need to be surveyed.

Using formulas to calculate a sample size

Yamane (1967) provides a simplified formula to calculate sample sizes. A 95% confidence level and p = 0.5 are
N
assumed for this Equation; n = 1+N e2 Where n is the sample size, N is the population size, and e is the level of
precision.

41