DATA ACQUISITION

Introduction
In this topic, we discuss the main sources of data, data collection techniques and measurement of
data, preparation of data collection instruments.

Learning Outcomes
On successful completion of this topic, you should be able to:
i. Define the term data
ii. Identify the various sources of data
iii. Describe different data gathering methods
iv. Explain the main data measurement scales
v. Explain data collection procedures

Nature and Sources of Data

Types of data

1. Quantitative – Numbers, tests, counting, measuring

2. Qualitative – Words, images, observations, conversations, photographs
Type of data to be collected can also be classified as follows

1. Secondary data- data someone else has collected

2. Primary data-data you collect
Internal and external sources of data

Quantitative and Qualitative

A lot of researchers or students in the social and natural sciences are always confronted with
gathering of information on the basis of surveys and experiments. Over the centuries researchers
have classified data into either qualitative or quantitative.

Information is qualitative when it is not countable or measurable. For instance, attitude of an

individual, colour, sex, qualification, etc., are good examples of qualitative information.
On the contrary, quantitative information is countable and measurable; this enables the
correct value(s) obtained.

Nowadays, researchers are going numerical in the sense that seemingly intangible concepts such
as intelligence, anxiety, etc. are now assigned numerical values. In the past, these intangible
concepts were seen as qualitative because expressed characteristics were merely described and
explained. There is the need to understand that the nature of data collected for particular study is
dependent on the following. These are; the purpose of research, population and sample of subject,
and the problem-focus of the research.

Primary Data Source

A lot of times, studies are geared towards achieving a specific purpose and these studies may
involve that the individual concerned creating his own data. At time, there are cases where there
are no data to assist the researcher or where the information available is not adequate for the
study; this may necessitate the generation of data for the purposes of the study. Primary sources
of data gathering are known to be connected with field observations, questionnaires, personal
interviews and experiments. These methods allow the data to be collected in line with need of the
study.

Secondary Data Source

They are data that had been collected and processed by an individual or organization and only
extracted for use by the present researcher. Such data are usually gotten from either unpublished
or published sources, such as books, journals, office reports and statistics, newspapers, magazines,
institutions records, government gazettes, National Bureau of Statistics (N.B.S), Central Bank of
Kenya (C.B.K), National Population Commission (N.P.C), Commission for University Education
(C.U.E), etc. Information from this source is already made; such that users have no control over it.
So, researchers are adviced to be cautious about their use. The reason is that at times these data
may be obsolete or may have been tampered with.
Internal Sources of Data
These are data that are generated within the confine of establishments themselves, in the course of
their day-to-day tasks. For instance, Business firms like those of financial institutions produce a
large financial data such as volume of transactions, balance sheets, data on income statement, data
on its staff, etc. higher institutions of learning generate and keep data on internally generated
income, enrolment figures per session, subventions from government, and a lot more. when an
organization works on its own created data and publishes same, which acts as its own source data,
such is called internal data.

External Data Source

They are data sources other than the internal sources. These are data that are utilized by
researcher(s) or organization(s) that do not generate them. They (external data) are numerous and
varied as they are gotten from published documents. External data source further is grouped into
primary and secondary sources.

Data Collection Methods/Approaches

Data collection is the process of gathering empirical evidence. The evidence collected helps the
study in getting new insights and thus answer the research problem. Selecting the data collection
method is one of the steps in the data collection process. Data collection consists of the following
steps:

i. Selecting methods of data collection

ii. Preparing and testing data collection instruments
iii. Identifying and selecting people to assist in data collection
iv. Training the research team in data collection
vi. Actual collection of data

Types/Approaches/Methods of Data Collection

There are a number of data collection methods. For any study, the appropriate method depends
On the research problem, design and the kinds of data required.
a) Questionnaire Approach: This approach is the commonest form of data collection in research.
questionnaire is the vehicle which aids data collection. It involves the distribution of question items
that are expected to produce important information from the targeted audience. These responses
are later transformed into computer readable form known as data through a process called
codification‘.
Three different means to administer questionnaire are:
(i) Questionnaires directly administered to the targeted audience by the researcher or the research
assistances.
(ii)Postal Questionnaires: The questionnaires are sent to the respondents via the mail, and the
respondents in turn return the questionnaires back to the researcher.
(iii) Questionnaires filled by the field officers who serve as interpreters in highly technical aspect
of the questionnaires, the field officers are permitted to help fill the questionnaires on-behalf of
the respondents because they know the questions better than they do.

What is a questionnaire?

• An instrument (form) to

– collect answers to questions

– collect factual data

• gathers information or measures

• A series of written questions/items in a fixed, rational order

A questionnaire is given to each respondent to fill on their own.

Advantages of questionnaires

• Can reach a large number of people relatively easily and economically

• Provide quantifiable answers

• Relatively easy to analyse

Disadvantages of questionnaires

• Provides only limited insight into problem

• Limited response allowed by questions
• May be not the right questions are asked
• Varying response
• Misunderstanding/misinterpretation
 • Hard to chase after missing data

(b) Observation Approach: this is the direct watching or noting of a phenomenon in an ongoing
event. In the non-experimental systems such as this, personal observation is a necessary. Data is
collected directly and used for research purpose.
This is a method, which is relevant to a research context where no direct response is required from
the subjects. It is a method appropriate to an observation design. Eye observation is made to: -

• Count the number of subjects

• Count of the frequency of occurrence of an activity/event
• Make an eye estimate

Advantages

• It is quick
• It is easy to carry out
• It is economical
• Doesn’t require personnel with high technical skills

Limitations

Not useful when detailed data is required

(c) Interview Approach: It is a method where the interviewer asks questions that are already
prepared in a questionnaire form and record the respondent‘s response in a space provided in the
questionnaire. The interview system is most appropriate when structured. The organized interview
has standardized questions that are formal and questions to all respondents must be precisely the
same and even arranged in the same way. This permits for uniform gathering of data from all the
respondents and as a result is easier compare to the unorganized interview.

(d) Report Approach: This technique of data gathering has a link with the secondary source of
data. It is a situation where already documented information is pooled together for a study purpose.
Data collected through the secondary sources fall under this category. These forms of data are
commonly used but could be misleading in some instances.

(e) Telephone Approach: It is another method of gathering information (data) for a particular
research. It is a process where questions are asked by the interviewer and the respondent responds
via a medium known as the telephone. In this case, the interviewer asks the respondents questions
and records them as approximately as possible. These responses are subsequently used to generate
the research data.

(f) Results of Experiments Approach: The experimental designs in research studies produce a1ot
of information. The information gotten from these experiments are regular happenings in
researches.

(g) Focus Group Discussions Approach: share many common features with less structured
interviews, but there is more to them than merely collecting similar data from many participants
at once. A focus group is a group discussion on a particular topic organised for research purposes.
This discussion is guided, monitored and recorded by a researcher (sometimes called a moderator
or facilitator). This method is becoming increasingly useful and important in research. This is
especially in Social research(e.g. community development, co-operatives, entrepreneurship, etc).
Because of the above and other reasons, individual members find it easier to comment on certain
issues under the support and security of groups. In addition, data collection in such situations
requires community leaders who can adequately represent the community.

It is unique in the following ways:

It is less structured, more informal and interactive and deeper in the method of enumeration.
It is relevant and suited to community-oriented research
It uses the representative social and leadership structures existing in the community or
population

It covers a whole community, which is a small population on its own merit.

It is used to collect both quantitative and/or qualitative data.
Census

If every number or item of the population is surveyed, this is known as a census. A Census is
conducted with definite objectives, which may be both general and specific. This method should
be selected only when there’s need to have information on every member in the population.
The method has the advantage of being comprehensive

Limitations

It requires a wide administrative organization

It needs a large number of personnel
Itis costly

It requires a lot of time

PREPARING DATA COLLECTION INSTRUMENTS

A data collection instrument is the tool used by the researcher to collect the required data from the
respondents. The most commonly used tools are:

• Interview guides

• Observation guide

• Questionnaire

Questionnaire Design Process

1. Specify the needed information. This requires you to review the research problem and review
the research questions/objectives/hypnosis. You also need to have a clear understanding of the
target population

2. Specify the interview method to be used – personal, telephone, or mail interview

3. Determine the content of the individual questions. Include only the questions that are

Relevant. Every question you ask should contribute to information needed or serve some specific
purpose.
Neutral questions should be placed at the beginning of the questionnaire to help in establishing a
rapport with the respondent.

Filter questions may also be used. e.g. a. What products do you sell?

b. If you sell product A, go to question no _

4. Design the questions to overcome the respondents’ inability to answer. They must be clear.

5. Decide on the question structure – open-ended (unstructured), closed-ended (structured)

or a combination of the two. Open-ended questions help respondents to express

Themselves and are very easy to formulate. However, they are the most difficult to analyse since
have to be coded in themes.

6. Determine the wording of the questions. A simple guideline for doing so is; Use simple
vocabulary

Avoid ambiguity

Avoid double-faced questions e.g. “Do you want to leave your current spouse and marry

Another one?”

7. Arrange the questions in the proper order (i.e. In the order you want them to be

Answered)

8. Identify the form and the layout of the questionnaire e.g. margins, spacing, fonts, leave enough
pace for one to write. For the closed questions, use the same forms. E.g. options arranged vertically
or horizontally. Number all questions and provide boxes for marking responses.

9. Reproduce your questionnaire

Enough copies for pre-testing

Pre-test for validity and reliability

Modify the questionnaire

Enough copies for the final study

Minimize Ambiguity

1. Ensure vocabulary is simple, direct, and familiar to all the respondents.

2. Don’t use technical jargon on non-technical people.

3. Avoid words that mean different things to different people

E.g. College, “volunteer”, “mkubwa”

4. Avoid words with vague or ambiguous meaning. Give the respondent an adequate frame of
time reference to interpret the question; else he will have difficulties in interpreting. You’ll also
find it difficult to analyze. E.g. Often Occasionally

5. Avoid words that will embarrass the respondent. e.g. “Have you stopped beating your wife?”
Use a disguised question instead

6. Avoid leading questions, e.g. “The girl-child is disadvantaged, don’t you agree?”

7. Avoid instructions that are potentially confusing, e.g. lengthy instructions.

8. Ensure that all the questions are applicable to all respondents; else ask filter questions e.g.

“Which credit program were you in before joining this one?”

Pre - Testing/Piloting the Questionnaire

A questionnaire is pre-tested to identify any deficiencies so that they can be corrected before the

main study. When pre-testing: -

• Be open-minded; open to criticism

• Select the pre-test sample carefully

• Ensure the sample size is appropriate – not too large, but with all the sample groups
represented.

• Use personal interviews

• Give clear instructions to the respondents

Actual Data Collection

If research assistants are to be used, they should be thoroughly trained. You should also get the
authority to carry out research (from NACOSTI). In addition, the following fieldwork guidelines
will be useful:

1. Asking questions – follow the order in which they appear, use the exact words in the
questionnaire, repeat questions that are not understood and probe the respondent where necessary.
2. Record responses as you do the interview to avoid forgetting

3. Use the respondent’s words as much as possibly

4. If using research assistants: - supervise them; ensure procedure of research is followed;

administer some questionnaires (as control); collect and edit their questionnaires daily; keep a
record of hours worked by each; control the assistants (follow sampling plan, daily records of areas
covered/interviews done, control cheating, get their opinion on the research, etc)

5. Ensure interviews are of good quality – appropriate introduction, precision when asking

Questions, probing, handling defiant/hostile respondents, terminating interview effectively.

6. Avoid data collection errors, which may arise out of suspicion, inaccurate responses.

These may be caused by the first impression, time pressure, fatigue, long questionnaire,

Fraud & deceits, etc.

Research Data and Its Measurement

Data are gathered using research tools in the course of undertaking a study. The graduations of
these tools are peculiar to the kind of research design embraced by the researcher. The graduation
of these tools is dependent on the measurement parameter and design of the tools. There are four
levels of measurement in research data. These are discussed shortly.

i). Nominal Level: Nominal level is the lowest level of measurements obtained in research studies.
It is a scaling measurement that assigns numerical values to attributes exhibit by human behaviour
under investigation. For instance, the sex of a human entity (if Male=1, Female=2 otherwise=0);
State of Origin (Nakuru=1, Mombasa=2 ...Kisumu=36); Marital status (married=1, single=2,
separated=3, divorced=4, widowed=5); etc. It should be noted that numerical names are just labels
for identification which cannot be ordered or added.

ii). Ordinal Level: Ordinal level has features similar to the nominal level but added to this is the
element of rank ordering in terms of highest to lowest or biggest to smallest. Ordinal level is a
scaling measurement which ranks the perception of an individual as regard issues under study. A
good example of ordinal level is the Likert scale use in questionnaires. Two types of Likert scales
exist. These are the five-point and the four-point scales.

Statement: Research method will not be helpful to me in the course of furthering my studies.
Response options:
. 5. Strongly Agree
. 4. Agree
. 3. Undecided
. 2. Disagree
. 1. Strongly Disagree
A case of 5-point Likert scale.
Statement: English is no longer useful to me in course of my studies.
Response option:
. 4. Strongly Agree
. 3. Agree
. 2. Disagree
. 1. Strongly Disagree
A case of 4-point Likert scale.
Ordinal scales cannot be added or subtracted.

iii). Interval Level:

This measurement scale (Equal-interval) also possesses the features of the earlier discussed data
measuring scales (nominal and ordinal levels). This in a way has made the interval scale a higher-
order scale which can be utilized in more sophisticated measurements. On an interval scale, the
distances are numerically equal and they also represent equal distances on the property being
measured. Unlike the ordinal scale, Interval scales can be added and subtracted. A good instance
of the use of interval scale is the measurement of temperature.

iii). Ratio Level: This version of data measurement scale (Ratio scale) is the highest level of data
measurement in research. This is so because, it has the features of nominal, ordinal and interval
scales. Some variables such as weight, area, speed, velocity, and many more are sets of data which
no other scale is appropriate except the Ratio scale. The ratio scale has an absolute or natural zero
(―0‖) which has realistic implication. When a measure is zero on a ratio scale, it indicates that the
variable has none of the things being measured. All the operations in mathematics are only possible
on a ratio scale because of the absolute zero (addition, subtraction. multiplication and division).
Ratio scales are better used in the laboratory. In the case of questionnaires where there is no
absence of opinion, Ratio scale will be useless.