Chapter Five

Sources and
Collection of Data

1
 Types and Sources of Data
 Data could be quantitative (expressed in numerical
form) or qualitative (expressed in the form of verbal
descriptions rather than numbers).
 Consider the following points when choosing
whether to collect quantitative or qualitative data:
 The purpose for which the data is required

 The subject matter

 The method of data collection

 The method of data presentation

2
 The purpose for which the data is required

 Quantitative data: to perform statistical analysis

 Qualitative data: providing a detailed impression of the

issue or characteristic concerned.

 The subject matter

 Some kinds of subject matter (e.g. Production, export

levels, prices, imports, income, etc.) are relatively easily

presented in numerical form.

 Others (e.g. Attitude to a new product, religious beliefs,

etc) more appropriately presented in qualitative form. 3

 The method of data collection
 The collection of quantitative data is based

on statistically designed survey procedures.

 The collection of qualitative data relies

primarily on detailed observation or

interview.

The method of data

presentation 4
Types and Sources of Data
 Depending on the source, the type of data collected could be
primary or secondary in nature.
i. Primary data: are collected afresh and for the first time
 original in character.
 Its advantage is its relevance to the user
 but it is also likely to be expensive in time and
money.
ii. Secondary data: are those which have already been
collected by someone else and passed through the
statistical process.
 It is information extracted from an existing source
 it is less expensive and time consuming to obtain 5
Cont…
 The selection of data sources depends on:
Purpose and scope of inquiry
Availability of time
Availability of finance
Accuracy/precision required
Methods of data collection

6
 Collection of Primary Data
 Primary data can be collected through experimentation or
surveys.
i. An Experiment:
 Examines the relationship between two factors by
manipulating one while measuring changes in the other.
 Two types of experiments:
 Field experiments and
 laboratory experiments

7
ii. Survey:
 Obtaining information concerning a phenomenon under study from
all or a selected number of respondents.
 The investigator examines those phenomena which exist in the
universe independent of his action
 Survey designs divided into:

– Cross-sectional design
– Longitudinal design

8
Survey Design

 A Cross-sectional survey
 Collects data at one time.
 Findings are generalized from such one-shot studies to the
sampled population only at the time of the survey.
 Is most appropriate if the aim of the research is single-time
description (political attitude, work motive)

9
 A longitudinal survey
 Takes place over time with two or more data
collections.
 Measuring changes overtime.
 Appropriate when the objective of the examination of
some dynamic process involving change over-time
 The following are the types of longitudinal surveys:
• trend survey
• cohort survey
• panel survey
10
A trend survey
 General population is studied over time.
 Used to study trends
 Different samples are selected from a general population
and studied at different time.
 While different persons or cases are sampled and studied in
each survey,
 Each sample represents the same population but at
different time.
 The members of the population will change overtime of
study.
E.g. Study on the attitude of workers of a proposed company-x
11
towards their management.
A cohort survey
 Specific population is studied over time.

 Is a survey design used in the study of process

 It focuses on the same specific population

each time data are collected.
 Data are collected from different samples that
are drawn from the same specific population.
 The same cohort or batch, at different points
in time
12
 A cohort survey

 Eg.1 Different samples could be selected at different time, at 5-years’

interval, from the student body graduating from a given university in

the class of 2000 to determine their attitude towards work.

 While the sample would be different each time, we would be describing

the class of 2000 (a specific population).

 If we study a class of 2005 the second time around, we would have a

trend study rather than cohort study.

 Eg.2: A cohort study of a given age group could be studied (say every

10-year).

 If we collect data on attitude of 20-years old youth towards marriage,

then we can do it on 30- years group the second round, 40-years

13
group the third round and so on.
A panel survey
 The same sample is measured two or more times.

 The samples can represent either a specific or a

general population.
 Indicate why the individual elements in the
population shows changes of behavior or attitude
 E.g. A researcher might interview all the members
of his panel (i.e. his “permanent” sample) at one-
month intervals throughout the campaign.
14
  Panel studies are difficult for the following reasons.

 Cannot be carried out through secondary analysis of previously collected

data.

 It is expensive and time consuming.

 The extent of non-response that occurs in later due to some of the

interviewees who are unable or unwilling to be interviewed.

 The analysis can be complicated especially when the study period is

relatively longer.

15
Types of Longitudinal Studies
 Trend studies: examine changes within
the general population over time.
 Cohort studies: examine more specific
populations as they change over time.
 Panel studies: same set of people are
interviewed two or more times over time.

16
 NB.

 Both trend study and cohort study can be accomplished

through secondary data analysis.

 Are more appropriate when the research problem deals with

broad trend overtime or process overtime.

 Panel study cannot be carried out through secondary analysis

of previously collected data

17
Selecting a method of data collection
 Units of Analysis

 understand what data applies to; as well as the

intended unit of analysis.

 E.g.1 If we are studying customers’ preference

to different shoe varieties, our unit of analysis

could be people’s buying behavior.

 The unit of analysis here are adult customers.

18
Cont…
 E.g.2 If we are studying the debt repaying
patterns among Small and Micro Enterprises
(SMEs), then our unit of analysis are not
individuals but firms (institutions).

 Data will be obtained from the Enterprises

(SMEs).

19
Measurement scales

 How to measure what we are interested in studying.

 Describe different types of variables having

different scales of measurement.

 Anticipate what type of analysis will be used before

selecting the variable in the study.

 Four types of variables/data (scales):

 Nominal Scale

 Ordinal Scale

 Interval Scale
20
Nominal Scale
 Divide responses into two or more distinct
categories.
 Each response is assigned to only one
category
 Responses differ only in kind, not in degree
or amount.

21
Nominal Scale
 E.g.1: Male/female, first - year/senior.

 E.g.2: Responses to the question: “Which

field would you like to specialize on?

 The two main response alternatives would

be Accounting or Finance.

 Recognize differences in the kind of

response alternatives

22
Ordinal Scale
 Further distinction by quantity of response alternatives.

 Show numerical differences between response

alternatives.

 Consider the following response alternatives to the

question,

 “What is your highest level of education?”

 (1) less than high school (2) high school graduate,

(4) college graduate, (5) beyond college.

 These response alternatives show increasing amounts of

23
Interval Data

 Interval data are similar to ordinal data in that

both reflect increases in quantity.
 With interval data the quantity/distance
between the different responses of the
variable is the same.
 A good example of an interval variable is
grade point average.

24
Interval Data

 A very popular type of scale in the social sciences is

the Likert scale .

 In Likert scale participants indicate their degree of

agreement with a stated attitude or judgment.

 Eg.: strongly agree, agree, neutral, disagree, and

strongly disagree.

25
Ratio Data
 With an absolute lower value that corresponds to the absence of

the measure.

 Have an absolute zero.

 E.g. consider an agricultural research study that has fruit

production as the dependent variable.

 A plant that produces zero pieces of fruit would receive a score of

zero — the absence of that entity.

 Allow numerical values to be placed in ratios.

 In our agricultural example, if Plot A produced an average of 40

pieces of fruit and Plot B produced an average of 20 pieces of fruit;

we could claim that Plot A produced twice as many pieces of fruit

26
 Survey Data Collection Methods

1. Observation Method
 Watching and recording behaviour and non-
behaviour characteristics of living beings,
objects or phenomena.
 Useful when studying subjects who are not
capable of giving verbal reports of their
feelings.
27
Advantages of observation method
 Subjective bias is eliminated, if observation is done accurately

 The information obtained relates to what is currently happening

 It is not complicated by either the past behavior or future intentions

 It is independent of respondents’ willingness to respond

 Less demanding of active cooperation

 Limitations
 Expensive

 The information obtained is limited

 Sometimes unforeseen factors may interfere with the

observational task. 28
 Types of observation

 Structured observation: the observer has a clear definition of the units to be

observed, the style of recording the observed information, the selection of the

appropriate data of observation, etc.

 Non-structured observation: the opposite of structured observation

 Depending on the nature of the observer observation is classified into three basic

forms:

 Secretive - where the subjects of the study are unaware that they are being

observed

 Non-participant: where the subjects of the study are aware that they are being

observed but the observer takes no part in the behavior being observed.
29
 Participant: when the subject and the observer interact.
2. Interview Method
 Involves oral questioning of respondents
either individually or as a group.

a. Personal interviews: - requires a person

(interviewer) asking questions in a face-
to-face contact to the interviewee.

b. Telephone interviews: -contacting

respondents on telephone itself.
30
Cont…
 Interviews can be: -

 Structured: - uses of a set of predetermined

questions and highly standardized
techniques of recording.

 Unstructured: - uses a list of topics rather

than fixed questions.

 Is more flexible and allows much greater

freedom to ask in case of need. 31
Advantages of personal interview
 More information and greater depth

 The interviewer by his own skill can overcome

respondents resistance

 Greater flexibility in case of unstructured interview

 Personal information can be obtained easily

 Sample can be controlled:

 No difficulty of missing returns

 Low non-response rate

 Language can be adopted to the ability of persons

interviewed
32
Weaknesses of personal interview

 Very expensive
 Possibility of bias of interviewer and
respondent
 Some respondents may not be easily
approachable (officials, executive bodies)
 Time consuming

33
Merits of telephone interview
 Faster than other methods

 Cheaper than personal interview

 Recall is easy

 Replies can be recorded without causing

embarrassment to respondents
 No field staff is required

34
Demerits of telephone interview

 Little time is given to respondents

 Survey are restricted to respondents who
have telephone facilities
 Not suitable when comprehensive answers
are required
 Questions have to be short and to the point

35
Preparing interview procedure
1. Determine general and specific research questions

 what do you intend to find out?

 what information is essential from the interview?

2. Draft the interview questions

 draft open questions

 draft closed questions.

3. Sequence the question

 group into topic sections

 vary question type

 arrange sections in sequence. 36

Cont…
4. Consider your process needs

 prepare suitable transitions;

 process questions.

5. Prepare the introduction and closing

 record verbal statements;

 note non-verbal statements.

6. Prepare for recording the responses

 decide on general method;

 organize procedure for written responses.

7. Pilot test the interview procedure

• pilot-test;
37
• revise as necessary.
3. Questionnaire
 Is a data collection tool in which written questions
are presented and given to respondents.
 It can be administered to respondents in different
ways:
– Sending questionnaires by mail.

– Gathering all or part of the respondents in one place at

one time.
– Hand-delivering questionnaires to respondents and
collecting them later.

38
Cont…
 Open-ended Questionnaire:

 Does not restrict the respondent in replay to the questions.

 Example: Why did you choose to do your graduate work at this

university?

_________________________________________________________________

__________________________________________________

 Close-ended Questionnaire

 Questionnaires that call for short, check-mark responses are

known as the restricted, or closed-form, type.

 Here mark a yes or no, write a short response, or check an item

39
 Example for close-ended:

 Why did you choose to do your graduate work at this university? Kindly

indicate three reasons in order of importance, using the number 1 for the

most important, 2 for the second most important, and 3 for the third most

important.
Rank
(a) Convenience of transportation ____
(b) Advice of a friend ____
(c) Reputation of institution ____
(d) Expense factor ____
(e) Scholarship aid ____
40
(f) Other, (kindly indicate)
 Cont…
 Rating Scale: involves qualitative description of a
limited number of aspects of a thing or of traits of
a person.
Strongly Agree Undecided Disagree Strongly
agree disagree

Superior Above Average Fair Inferior

average

Excellent Very good Average Below Poor

average

Always Frequently Occasionally Rarely Never 41

Example

 The private Banks give sufficient interest rate that attract

citizens to open saving accounts. (Please, indicate your level of
agreement or disagreement by circling the corresponding scale)

 Strongly agree = 5; Agree = 4; Undecided = 3;

Disagree = 2; Strongly disagree = 1

How to improve questionnaire items or useful tips

to improve questionnaire (Read your material)
42
Essentials of a good questionnaire
 Should be short and simple

 Should proceed in logical sequence

 From easy to more difficult questions

 Personal questions should be left to the

end
 Unclear expression capable of different
interpretations should be avoided
43
  Merits of Questionnaire

 Is free from the bias of the interviewer; answers are in respondents’ own words

 Respondents have adequate time to give well thought out answers

 Respondents who are not easily approachable can also be reached conveniently

 Demerits of Questionnaire

 Can be used only when respondents are educated and cooperating

 The control over questionnaire may be lost once it is sent

 There is inflexibility because of the difficulty of amending the approach once

questionnaires have been dispatched

 There is possibility of unclear replies or omission of replies altogether to

certain questions 44
4. Focus Group Discussion/FGD/ Method
 Is typically composed of 7 to 12 participants who are
unfamiliar each other and conducted by a trained
facilitator.

 Eight is a popular size.

 The facilitator sets the agenda and lets the participants

brainstorm on the agenda.

 The participants should be more or less homogenous

45
Cont…
i) Provide explanations for the way in which survey

respondents have responded.

ii) Can be used to check on and verify data obtained

through other methods

iii) Can serve as convenient means of obtaining data on a

wide population within the shortest time and economy

of cost.
46
Case Study Method
 Is an intensive study of a case which
may be an individual, institutions, a
system, a community, an organization,
an event, or culture.
 Is a comprehensive investigation of
the particular unit under consideration

47
Characteristics of case study method

 A researcher can take one single social unit or more of such units

for his study purpose

 The selected unit is studied intensively

 A researcher has to make complete study of the social unit

covering all aspects

 The effort is to know the mutual interrelationship of causal

factors

 The behavioral pattern of the concerning unit is studied directly

48
and not by an indirect and abstract approach
Collection of Secondary Data
 Sources of secondary data can be:

 Central Statistical Authority Publications

 Publications by Regional Governments

 Publications by the different Ministries

 Publications of the National Bank of Ethiopia

 On-line and Electronic Data Bases

 Reports and publications of various

associations
49
 Cont…
 Before using secondary data, one should verify: -

 Reliability of data: - by answering questions like:

 who collected them?

 What were the sources of data?

 What methods were used?

 At what time were they collected?

 How they were analyzed?

 Suitability of data: Data must be evaluated whether they

could serve for another purpose other than the one for
which they were collected.
 Adequacy of data: This should be done in terms of area
50
 Factors to be considered when selecting
appropriate method for data collection :
 Nature, scope and object of inquiry:

 In deciding whether the data already available (secondary data) are to

be used or the data not yet available (primary data) are to be collected.

 Availability of funds: When funds at the disposal of the researcher are

very limited, select a comparatively cheaper method

 Time factor: Some methods take relatively more time, whereas with

others the data can be collected in a comparatively shorter duration.

 Precision required: Considered at the time of selecting the method of

collection of data.

51
The End!

52