Chapter 7: Data Collection


[Chapter 6] [Back to Research Guide][Chapter 8]

7.1 Introduction

Data collection is one of the most important stage in conducting a research. You can have the best research design in theworld but if you cannot collect the required data you will be not be able to complete your project. Data collection is a very demanding job which need thorough planning , hard work, patience, perseverance and more to be able to complete the task successfully. Data collection starts with determining what kind of data required followed by the selection of  a sample from a certain population. After that, you need to use a certain instrument to collect the data from the selected sample.

7.2 Types of Data

Data can be divided into two types, namely quantitative and qualitative. Quantitative data is numerical in nature and can be mathematically computed . Quantitative data measure uses different scales, which can be classified as nominal scale, ordinal scale, interval scale and ratio scale. Nominal scale is used  to categorize items into different group, for example male and female, different examination grades, different races in a country, different types of companies, different shoe sizes  and so on. The statistics that is used to analyze this type of data is mode. The second measure for quantitative data is the ordinal scale which other than providing information as the nominal scale does also permit ranking of the data. For example, shoe sizes can be ranked from big to small. A typical ordinal in ranking a set of data is poor, satisfactory, good, very god, excellent. On the other hand, an interval scale not only rank order of a set of data, it also measure the order in units of equal intervals. However, the starting point for the measurement is arbitrary rather than absolute. For example, Celsius in temperature measurement is an interval scale because 0o C is an arbitrary value, it is not absolute zero, it does not mean temperature is lacking at that point. For absolute quantities, we have to use the ratio scale. For example, measuring temperature in Kelvin is a ratio scale as it starts the measurement at absolute zero, which is about -273o Celsius.

Qualitative data are mostly non-numerical and usually descriptive or nominal in nature. This means the data collected are in the form of words and sentences.

7.3 Sources of data

Generally we can collect data from two sources, primary sources and secondary sources. Data collected from primary sources are known as primary data and data collected from secondary sources are called secondary data.

Primary data are also known as raw data. Data are collected from the original source in a controlled or an uncontrolled environment. Example of a controlled environment are experimental research where certain variables are being controlled by the researcher. On the other hand, data collected through observation or questionnaire survey in a natural setting are examples data obtained in an uncontrolled environment. Secondary data are data obtained from secondary sources such as reports, books, journals, documents, magazines, the web and more.

7.4 Data Collection Methods

There are many methods to collect data, depending on our research design and the methodologies employed. Some of the common methods are questionnaires , interview and observation.

7.4.1 Questionnaires Survey

According to Wikipedia, a questionnaire is a research instrument consisting of a series of questions and other prompts for the purpose of gathering information from respondents. Although they are often designed for statistical analysis of the responses, this is not always the case. The questionnaire was invented by Sir Francis Galton. Questionnaires are widely used for both quantitative and qualitative research.

Questionnaires often use various measuring scales to obtain information from the respondents. To obtain the bio data of the respondents, we use nominal scale as it serves as label or identification such as gender and age, which often does not involve calculations. To gather respondents' preferences, we use ordinal scale which is used to arrange objects or alternatives according to their magnitude in an ordered relationship. To obtain information related to attitude, we use rating scales. Rating asks the respondents to estimate the magnitude of a characteristic or quality regarding certain object or thing.

Questionnaire is designed both for descriptive as well as analytical surveys. In a descriptive survey, the questionnaire will normally use nominal and ordinal scales because it concerns primarily with the particular characteristics of a specific population of subjects. It does not required the examination of dependent and independent variables. Examples of questions asked in a descriptive survey are shown below:

  • State the location of your company
  • How many workers are employed by your company?
  • When was the company founded?

On the other hand, rating scale is always used to measure attitude or opinion of the respondents in an analytical survey. In an analytical survey, it normally needs to identify independent, dependent and extraneous variables based on a certain conceptual framework. A researcher needs to conduct a thorough literature review by paying attention to any existing research and theory relevant to the research problem. Once the variables are determined, they are built into a questionnaire using rating scale measures, the most popular one being the Likert scale. Likert scale questions can consist of three items, four items, five item, six items, seven items and more. The ratings obtained from the respondents are then summarized or averaged up to reflect a certain variable, such as job satisfaction. They can be used to analysed any causal relationship with other variables. Example of Likert-type questions are shown below:

                                                                                                       Strongly Disagree                    Strongly Agree   

  • . Need additional staff to manage electronic commerce          1               2              3             4               5       

  •    Difficult to justify the cost with desired benefits                    1               2              3             4               5

  •    Information from the electronic commerce is not useful        1               2               3             4              5


When we are designing a questionnaire, we have to pay attention to the following issues:

  • Are the instructions clear and unambiguous?
  • Can the questions be understood; are they free from jargon, esoteric terminology, unsuitable assumptions and ambiguity?
  • Are the respondents posses the requisite information and knowledge to answer the qusetions?
  • Is the wording of questions appeared offensive and embarrassing to the respondents?
  • Is the wording of questions lead to bias through leading the respondent to particular answers?

Questionnaire can be administered by post, face to face distribution of the questionnaires or by interviewing the respondents. If you are conducting questionnaire survey by interview, you have to keep interviewer bias to the minimum by following certain rules such as

  • Record exactly what the respondent answers
  • Do not answer on behalf of the respondent
  • Read the questions clearly and slowly.
  • Do not show approval or disapproval of any answer

7.4.2  Observation

Observation is a process of recording the behaviour patterns of people, objects, and occurrences without questioning or communicating with them. Observation can take the place in a laboratory setting or in a natural setting. Generally there are two ways to conduct observation, namely non-participative observation and participative observation.

The researcher in non-participative observation does not involve in the activities of the people being observed. He or she merely record whatever happens among the people , including their actions and their behaviour, and anything worth recording. On the one hand, the researcher in a participative observation involves fully with the people being observed, with the objective of trying to understand the values, motives and practices of those being researched.

The main advantage of observation as compared to questionnaire survey is you can obtain richer and more in-depth information. You can able to catch phenomena, characteristics, activities and other things impossible to detect by questionnaire survey . However,there are some weaknesses associated with observation method as shown below:

  • Cannot control variables in the natural setting
  • Researcher own values and ethics might affect his objectivity and give rise to observer bias
  • Failure to observe some activities due to distractions.

[Chapter 6] [Back to Research Guide][Chapter 8]