Primary sources- Questionnaires

Research techniques that can be used to generate primary data include questionnaires, structured and unstructured interviews, formal and informal observations, measurements and scientific experiments.

Questionnaires

Research in health and social care often relies on obtaining information from individuals using social research methodologies. Questionnaires are convenient for obtaining information from many individuals in a survey and those answering them are sometimes known as respondents. A questionnaire asks questions of participants but, usually, also only enables answers to be given according to options constructed in advance by the researcher using a response frame. The response frame usually only allows a limited choice of different responses.

In small surveys, questions may be put to the respondent orally by the researcher, who then ticks a box corresponding to the answer given and also records the responses provided. Market research carried out in the street usually follows this approach.

In health and social care, this method may be the only way to capture information from a frail older person about the care they receive, for example.

In a self-completion questionnaire, respondents fill in the answers themselves. The questionnaire may be distributed to individuals directly by hand, by post or online, provided the organisation or researcher has a means of making contact through an email address or information held on a database.

You may well have been asked to complete a customer satisfaction questionnaire but being asked to complete a questionnaire is not the same as actually doing so. The respondent may ignore it, only answer some of the questions or not return the questionnaire to the researcher. The response rate to a questionnaire is an important measure of how representative the responses actually received by the researcher are, compared to the number of individuals originally asked to complete the questionnaire.

The sample population should be defined when the research is planned, as it will influence what information can be obtained from a survey. The method for selecting the participant sample from the sample population should also be considered at the planning stage.

Some frequently used sampling techniques include:

• random sampling, in which individuals are selected randomly; researchers may use random numbers selected by a computer or from a table of random numbers

• systematic sampling, which involves selection of the individual at a regular interval, e.g. distributing a questionnaire to, say, every sixth student who enters the college canteen, or every other patient who presents with diabetes

• quota sampling, which requires the researcher to select a pre-determined number of individuals from representative groups (e.g. according to age, area of the country, socio-economic profile, male and/or female etc.); opinion pollsters use this method

• opportunity sampling, which involves researchers handing out questionnaires to individuals who happen to be passing by at the time. Standing in a college canteen on a particular day and handing out questionnaires to anyone who will take a copy would be opportunity sampling.

The sampling technique and sample size chosen for a research study affect the validity of the research and the conclusions that may be drawn from it.

Who your respondents are could affect the interpretation of the results so it is usual to gather factual information about them that your secondary research has suggested might be significant.

Questionnaires usually request information on age, or age group and gender. However, requesting information that is not relevant to the research (e.g. about marital status) would be unethical because it invades individuals’ privacy. Ethical approval may involve removing some questions if this were the case.

The simplest questionnaires have response frames that offer only straightforward choices and require all responses to be indicated entirely by ticks. More complex questionnaires may have:

• more questions

• more options in each response frame

• require respondents to provide written comments.

Questionnaires with only closed and open questions are probably the simplest response frame for the researcher to construct but provide more limited data, which can be more difficult to interpret, especially when working with a small sample.

Closed questions only offer two alternative answers: usually ‘Yes’ or ‘No’.

Open questions require the respondent to answer freely, either orally or by writing in a blank space provided on the questionnaire. The respondent has to write the answer (or the researcher, if it is not a selfcompletion questionnaire).

Other response frames enable more specific and detailed information to be obtained. Examples include:

• ranking scales, which require respondents to rank different statements in an order, often using a number scale, where, for example ‘1’ is very important and ‘5’ is not at all important

• Likert scale response frames, which gather respondents’ opinions of carefully worded statements using a five-point scale such as ‘strongly agree, agree, neither agree nor disagree, disagree, strongly disagree’. Other descriptor words may be used.

Constructing a questionnaire takes time if it is to yield good quality research information. Questionnaires are very useful when finding out about people’s opinions, perceptions, experiences, or to find out how much knowledge and understanding they have of a topic.

Surveys undertaken for a student project are likely to focus on this type of information because of the ethical constraints on students regarding research in health or social care settings or with users of services.

The drawback of using response frames is that the answer options offered may not include the answer the respondent thinks is right for them. One way round this is to include an option such as ‘none of these’, ‘all of these’ or ‘other’, with a space for the respondent to provide an alternative answer.

Partly completed questionnaires reduce the quality of the data from a survey and therefore limit the validity of any interpretation and conclusions drawn from the research. Factors to consider when designing a questionnaire include:

• a clear understanding of the contribution the data respondents will make to the research

• the abilities and experience of the respondents, e.g. their understanding (avoid jargon), literacy skills, etc.

• how you will address ethical issues (e.g. providing information about the research)

• what the document looks like; does its layout, font size and style help the respondent complete the questionnaire?

• what instructions the respondent will need to complete the questionnaire

• the order of the questions, e.g. simple questions at the start and more complex ones later; questions probing personally sensitive information are best placed towards the end of the questionnaire

• how you will hold the interest of the respondent so they answer all the questions

• the distribution method and how long the questionnaire will take to complete, e.g. will respondents have a hard surface available for the writing involved?

• how the completed questionnaires will be returned.

Distributing questionnaires by email or post means that you need the email or postal addresses of the participants. This information is confidential and may not be known by the researchers. All such information would be subject to the DPA. The aim of any survey is to get as many of the completed questionnaires returned as possible; the return rate is an indicator of the reliability of the data generated from the survey.

To calculate the return rate, you need to record exactly how many questionnaires are printed and distributed, as well as how many are returned.

Activity

Produce a table outlining the strengths and weaknesses of using questionnaires to gather data. Include examples of when you should or shouldn't use questionnaires to gather information in different research areas (e.g. not suitable for a chemical reaction, suitable for gathering opinions on something)

Last modified: Monday, 30 Mar 2020, 14:20