12.4 Collecting data

Different ways of collecting data (EMG6W)

The aim of the research influences the way data will be collected. Four methods of collecting data are:

1. Observation: Data collection using observation does not entail personal contact. Counting the number of vehicles crossing an intersection every hour would be a good example of observational data gathering.

2. Interview: This takes place usually between two people where one is called the interviewer and the other is the interviewee or respondent. This method is usually chosen when it is convenient to talk to the respondents directly. For example if we wanted to determine whether people were happy with the way they were treated by sales staff.

3. Questionnaire: a questionnaire is a predetermined set of questions given to a number of respondents to complete. This instrument is good for getting information from many people. Questionnaires are also appropriate for getting information from people that are spread over a wide area and that are not easy to contact face-to-face. A questionnaire should have a short explanation of what your research is about. As with all data collection methods, questionnaires should always adhere to ethical and moral codes of conduct. An example of a questionnaire in use is the national population census for South Africa, which takes place every ten years (the last one was in 2011).

   

4. Databases: Sometimes we can use information that is already stored in a database, so that we don't actually have to find the data. Databases are simply organised lists of data - the list of learners at your school is a kind of database. Databases can be computerised, books or paper filing systems. A big advantage of these is that the data is already organised and is easy to access.

The method of collecting data must be suitable for the type of research we are doing. Let's look at examples to see why.

Deciding who to ask (EMG6X)

The group that we want to collect data from is called the population. In some cases, we can ask every person in the group we are interested in, to answer a questionnaire. Of course, not everyone will answer. The higher the number of respondents you get, the more valid your data will be.

In other cases, we need to choose a sample of people from the population. The choice of sample can have an effect on the reliability of the data and could even lead to sample bias.

Sample

A small selection of a larger population or collection.

Sample bias occurs when a certain section of the population from which the sample is drawn is not representative of that population. You, as the researcher, need to find a way to take a sample that is likely to represent the population well. For example, if you want to find out what learners at your school think about physical exercise and keeping fit, the soccer team would not be a good sample. They don't represent the rest of the learners, and might be biased, because they are more likely to be fit and enjoy exercising. You should rather find of a way of giving every learner an equal chance of being in your sample. One way would be to ask every tenth learner who arrives at the school gate in the morning.

One way to avoid sample bias is to select a random sample. A sample is random if every member of the population has the same chance of being selected - the interviewer doesn't choose particular people. Asking every tenth learner who arrives at the school gate would be an example of a random sample. Sometimes random samples can still result in sample bias however - so it is always important that your random sample is still representative!

How to develop a good questionnaire (EMG6Y)

The questionnaire also has an important role in making sure that the information you collect is valid. You should aim to get a high number of respondents and accurate information. If not enough people fill in the questionnaire, then you don't know whether the information you get reflects the real situation.

The tips below help you to make sure that your questionnaire is clear and accurate, and also that people are likely to complete it.

1. Keep it short. Don't include information that you already know.
2. Write down all the relevant questions you can think of. Next, analyse the appropriateness of each question by asking these things:
   • Is this question necessary? If not, don't include it in the questionnaire.
   • Is it possible for the respondent to answer this question? Don't assume that the respondent can remember something that happened five years ago, or that he or she will have certain information.
   • Will the respondent answer the question honestly?
   • Can the question be answered quickly?

You can make some questions easier to answer. Using categories instead of precise answers may also make it easier for respondents to complete the questionnaire. For example, most people do not like telling people their age, weight or salary, but grouping those numbers into categories makes it easier to answer.

1. Decide how to ask the question. There are two different types of responses: open-ended and closed-ended.

   In an open-ended question, the answer is usually the opinion of the respondent and the respondent can answer in their own words. In this way you can gain insightful data and avoid receiving answers that are biased. A disadvantage to this type of question is that respondents might leave it out if it takes too long to answer.

   Closed-ended questions could give respondents some options for the respondent to choose from, which is convenient because they can simply tick the right box.

2. Check the wording for each question. Look at the questions you have written and ask, “Will people be able to understand this question?”
3. Decide on an order of questions that is easy to understand.