Instruments of data collection are tools and materials with which you collect data for your research. There are a variety of instruments in research parlance. Here we’ll look at a number of them. Note that the instrument you adopt in collecting data is solely dependent on the type of research you’re undertaking.

Data collection instruments in research are also called measuring instruments, these instruments include; questionnaires, field observation, oral interviews, archival documents, laboratory experiments and scales as explained below:

Questionnaires: Involves collecting information from respondents through well-structured questions.

Field Observations: Involves visitation to study areas to gather relevant data on the recent happenings regarding the subject matter under study.

Oral Interviews: Involves collecting information from respondents through personal interviews. The interview can be one-on-one or via a telephone call.

Archival Documents: Involves collecting relevant information from government records or stored materials. You can equally collect data from the National Bureau of Statistics or any related establishments.

Laboratory Experiments: Data collected from laboratory results can be classified under data source.

Scales: This involves data or results from measurements obtained. They can form part of data sources.

What is Data Collection?

Data collection is collecting or gathering relevant data or information from an area, respondents or event to measure and analyze the collected data using accurate measuring instruments and other data collection tools in research. Data can be collected through various methods.

The methods of data collection in research are the ways or processes by which the data are collected for research. Data collection methods include experiments, direct observation, questionnaires, survey approach, interviews, tests, review of existing literature, and data sourced from archives.

Example of Data Collection in Research

An example of a Questionnaire as a data collection instrument is shown below.

Instruction: Here are some statements about Biological Science. You respond by ticking in the box, which mostly represents your feelings. There are no wrong or right feelings. We only want to know how you feel about the study of Biology to improve your pass rate.

Personal Profile of Respondents:

Name of school you teach?…………………..

What subject do you teach?…………………..

What’s your gender? ……… male (), female ()

Question 1.

Why do you opt for Biology?

1. It is a requirement for my career ()
2. To complete my enrollment requirement ()
3. We were told to offer at least one science subject ()
4. None of the above

Question 2.

Kindly tick your most preferred opinion from the list of options in the table below.

Importance of Data Collection

Data collection helps the researcher or investigator with answers to the research questions, helps the researchers make informed decisions about the subject being researched, and helps the researcher find solutions to the problems under investigation, data collected also improves a research argument and beefs the investigator’s opinion and conclusion about a study. When correct data are gathered, they help produce reliable analysis which contributes to the researcher’s recommendation at the end of a study. The outcome from data collection also validates the research outcome and improves and produces a reliable and accurate study.

What Are the Primary and Secondary Data Collection?

Data can be sourced through two major means namely; the primary and secondary sources of data collection. These are explained below.

Primary Data Collection

Primary data are data obtained through the administration of questionnaires, one-on-one interviews and direct observations from the study area.

Secondary Data Collection

Secondary data are sourced from second-hand materials, information or data collected from published or written materials by other authors in the form of textbooks, newspapers, reports, magazines, publications and journals.

Secondary data already existed before the research was undertaken. Secondary data can further be sourced from annual reports and accounts of organizations and the Federal Office of Statistics etc.

Important:

All sources as stated above may not be used at the same time. Each of them has a purpose they serve and is adopted based on the researcher’s discipline and the type of research carried out. Moreover, 3 out of the 6 sources listed above are mostly used and are instrumental for collecting valid data. These 3 are:

Questionnaires (Widely used in both qualitative and quantitative research), Interviews (Suitable for qualitative research), and Observation (Suitable for both qualitative and quantitative research).

The questionnaire being the most used instrument for collecting data in research will now be explained in detail.

Questionnaire Design in Research

The questionnaire is a very important instrument for data collection in a survey. Data are collected according to the survey design. Below is the procedural format for designing a questionnaire.

1. Design the survey:

The design may involve grouping the subjects of the population into two or more groups. The design may also classify the respondents. The item for grouping the population subjects must be established. Also, items for classifying respondents should be established.

2. Operationalize the aspects of the independent variables investigated and the dependent variable of the study. The operationalizing questions will be used to construct the questionnaire.
3. Decide on an appropriate scale for presenting the response options.
4. Write the cover letter
5. Prepare the questionnaire

Questionnaire Scales

There are various scales for presenting the response options. But there is a caveat;

Use a scale that the respondent is familiar with, use a scale that you can translate the responses to data for analysis and use a scale that yields data which you can analyse and interpret.

Nominal Scale:

This is the categorical scale that measures a concept at two points. The two points may be labelled “Yes” and “No” or “Agree” and “Disagree”.

The respondents are very familiar with this scale and it is very easy to translate the responses to data. The scale is a qualitative scale and the data generated from it are qualitative data.

On a nominal scale, we translate the responses to data by carrying out frequency counts. The sizes of the frequencies are often expressed as percentages. To illustrate the translation process, consider the following measurement instrument;

Assume the instrument (that is, the questionnaire in this case) is administered to 50 respondents. For each item, we count the respondents that answered “Yes” and those that answered “No”. The number of respondents answering “Yes” gives the frequency for “Yes” and similarly for “No”.

All measurement scales must provide for how statements should be classified. The usual principle is to classify a statement using the response options, unless, other criteria are indicated.

Nominal data are interpreted using the percentage. For example, the data analysis for the table below would run thus;

70 per cent of the consumers (assuming that the respondents are consumers) are of the view that Nigerian manufacturers are not careful in the design of their products. Only 30 per cent admit carefulness and meticulousness in product design. The findings are stated from the interpretation of nominal data using the modal response to each statement.

Likert Scale:

It is a quantitative scale that measures a concept at 5points on a continuum, which are labelled as follows:

Uncertain or undecided or neutral

Strongly disagree or disagree to a large extent

Disagree or disagree to a limited extent

Agree or agree to a limited extent

Strongly agree or agree to a large extent

To translate the responses to data for interpretation, the options are scored as follows:

Uncertain = 0

Strongly disagree = 1

Disagree = 2

Agree = 3

Strongly agree = 4

With these scores, a mean score is computed for each statement. To illustrate, assume that the questionnaire is administered to 50 respondents. The computational procedure is summarized below;

The arrangement shows that out of 50 respondents, 12 were undecided on the issue of whether producers are careful in product design, 18 strongly disagree, 10 disagree, 8 agree, and 2 strongly agree. These are the frequencies. When the frequency of an option is multiplied by the score of the option, a weighted score for that option is obtained. Thus, the weighted score for the disagree option is 20×10=20. The total weighted score for all the options is divided by the number of respondents to obtain a mean score for each statement;

Weighted mean score = 70/50 = 1.4

Data generated by the Likert Scale are interpreted concerning the options via the scores. For example, a mean score of 1.4 lies between strongly disagree to disagree. It is closer to strongly disagreeing than agreeing. The same observation pattern shows that a mean score of 1.6 is closer to disagree than strongly agree. In the same vein, 1.2 lies very close to strongly agreeing.

1. Semantic Difference Scale

This is also a quantitative scale that measures a concept on a common, which ranges from 1 to 7. The respondents are asked to rate whatever is being measured by ticking a score via the responding box.

The inbuilt scores in the scale may be interpreted as follows:

1=very low rate

2=low rate

3=slightly below average

4=average

5=slightly above average

6= high

7=very high

The procedure for generating data from a semantic differential scale is similar to that of the Likert scale where an average score is also derived.

How to Write A Questionnaire Cover Letter

It is necessary to write an introductory note, which is known as the cover letter, to the questionnaire. In the cover letter, you may briefly introduce the problem and the major aim or you may hide it. If the problem and aim are stated on the cover letter, the questionnaire is said to be undisguised otherwise it is disguised. The cover letter must direct the respondents on how to respond. If a semantic differential scale is used, the interpretation of the score should be stated in the cover letter.

In general, the cover letter should be fine-tuned to motivate the subject to respond. Expressions such as “I am a postgraduate student…” should be avoided. On the other hand, expressions such as “you are one of the most carefully selected persons for this study” may be added to swell the heads of the respondents. You need not promise what you cannot do; for example, promising a copy of the research findings (a synopsis) upon completion of the study. This is not necessary. You don’t even need to mention that “this research is being carried out under the auspices of…department of …, university of…” The respondent any currently be facing threats from above.

The cover letter should also contain the retrieval date of the questionnaire to remind the respondents of when you will be coming to collect back the questionnaire.

Questionnaire Design Illustration

To illustrate how to construct a Questionnaire, we shall consider an example.

Problem: The performance of students in Biology examinations is relatively poor. Biology teachers are worried about the situation and call for an investigation.

Issues for investigation: Attitude and personality.

Aspects of the Issues Investigated:

1. The attitude of the student to the learning of Biology
2. The attitude of the teacher towards the teaching of Biology
3. The teacher’s knowledge of Biology
4. Anxiety
5. Open-mindedness

Example of Questionnaire as a Data Collection Tool

Instruction: Here are some statements about Biological Science. You respond by ticking in the box, which mostly represents your feelings. There are no wrong or right feelings. We only want to know how you feel about the study of Biology to improve your pass rate.

Questions.       

Why do you opt for Biology?

1. It is a requirement for my career ( )
2. To complete my enrollment requirement ( )
3. We were told to offer at least one science subject ( )
4. None of the above ( )

Oral Interview Schedule in Research

The investigator can use open-ended questions if he wants the respondents to substantiate their response to a particular item in the questionnaire. Alternatively, construct an oral interview schedule and administer it to get oral information for substantiating the response.

Oral interviews should be used to collect data for the study if the study is designed to seek the knowledge of (insight) of practitioners (experts). Before any systematic attempt is made to collect the insights of experienced practitioners, the investigator must have preliminary ideas of the important issues in the area. The preliminary data usually comes from the literature review. The ideas are then used to construct an interview schedule. To concretize the preliminary ideas, you should pre-test the interview schedule. The response in the response in the pre-test should not be included in the final administration. Before administering the final draft, it is important to make a pre-contact through which a copy of the interview schedule is given to the respondent to prepare him or her for the interview.

Sorting the Questionnaire on Retrieval

The retrieval process of the questionnaire may not be easy. You may have to do a follow-up. Worry the respondents to make sure you get their response but be patient. Always go with an extra questionnaire so that if there is an excuse for loss, you can immediately replace it.

When the questionnaire is retrieved, sort it out according to the groups in the study. Certainly, there must be an item in the questionnaire that requests the respondents to indicate the group they belong. It is that item that is used to sort the questionnaire into groups. It is after the questionnaire sorting that the data are extracted, separated for each category and analyzed.

Examples of How to Write the Instruments Used to Collect Data in Research