Sociology Research Methods: Practice Questions

These are original questions written for Marlbridge, in the style and at the standard of the examination. They are not reproduced past-paper questions — examination boards hold copyright in their own papers. Use these alongside the official past papers available free from your board.

Related: Research Methods revision notes

---

Section A

1. Explain the difference between a research method and a research design. [2]

2. State four factors a sociologist must consider when choosing a method. [4]

Section B

3. Explain two strengths and two limitations of participant observation. [8]

4. Explain two strengths and two limitations of using official statistics as a source of data. [8]

5. Explain the difference between reliability and validity, and explain why a method that is high in one is often low in the other. [6]

6. Explain three ethical principles a sociologist must observe, and explain the ethical difficulty of covert participant observation. [9]

7. A researcher wishes to study a group of homeless people.

(a) Explain why random sampling would be impractical. [2] (b) Suggest a suitable sampling method and justify it. [3]

---

Answers

1. A method is the specific technique used to collect data, such as a questionnaire or an interview [1]. A design is the overall structure of the study — who will be studied, in what conditions, and in what sequence [1].

2. Any four: practical factors — time, cost, access and the researcher’s skills [1]; ethical factors — consent, harm and confidentiality [1]; theoretical position — positivist or interpretivist [1]; the nature of the topic and the group being studied, particularly whether it is sensitive or hard to reach [1].

3. Strengths: it produces highly valid, in-depth qualitative data, because the researcher sees what people actually do rather than what they say they do [1] [1]. It allows the researcher to achieve verstehen — an understanding of the meanings the group attaches to its own behaviour — which no questionnaire could capture [1] [1]. It is also useful for studying groups who would not agree to be interviewed [1]. Limitations: it is very time-consuming and expensive, and the sample is necessarily small and unrepresentative, so findings cannot be generalised [1] [1]. It is unreliable — the study depends on the personal relationships and interpretations of one researcher, so it cannot be replicated and there is a risk of “going native”, losing objectivity by over-identifying with the group [1] [1].

4. Strengths: they are readily available, free and cover very large populations, often nationally and over many years, so trends and comparisons are possible at no cost to the researcher [1] [1]. They are usually collected in a standardised way, making them reliable and quantitative, which suits positivist research [1] [1]. Limitations: they are collected by the state for its own administrative purposes, not for the sociologist’s research question, so the definitions and categories may not fit — the definition of unemployment, for example, has been changed many times [1] [1]. They may lack validity: crime statistics record only offences reported to and recorded by the police, so the “dark figure” of unreported crime is missing entirely [1] [1].

5. Reliability means a method is consistent — repeating it would produce the same results [1]. Validity means the data gives a true and authentic picture of what is being studied [1]. They often pull in opposite directions because standardisation creates reliability but restricts what respondents can say [1] — a structured questionnaire can be repeated exactly, but forces answers into pre-set boxes that may not reflect people’s real views [1]. Conversely, unstructured methods let people express themselves freely, giving validity, but are shaped by the individual researcher and cannot be repeated identically [1] [1].

6. Any three, 2 marks each: Informed consent — participants must know what the research involves and agree to take part, with the right to withdraw [1] [1]. Confidentiality and anonymity — identities and data must be protected, using pseudonyms and secure storage [1] [1]. Protection from harm — participants should not be exposed to physical or psychological harm, or to legal risk, greater than in everyday life [1] [1]. Covert participant observation is ethically difficult because participants cannot give informed consent, since the whole method depends on deceiving them [1]; the researcher may also witness or be drawn into illegal or harmful acts without being able to intervene or report them, and the group’s trust is exploited [1] [1].

7. (a) Random sampling requires a complete sampling frame — a list of everyone in the population [1]; no such list of homeless people exists, since by definition they have no fixed address to be registered at [1]. (b) Snowball sampling [1] — the researcher contacts one or two individuals, perhaps through a shelter or outreach worker, and asks them to introduce others [1]. This works because it uses existing trust networks to reach a hard-to-access group, though the sample will not be representative and members of the network may be similar to one another [1].

8. Strength: an experiment is controlled, so variables other than the one being tested can be held constant, allowing the researcher to identify cause and effect with more confidence than other methods allow [2]. Limitation: the controlled, artificial setting means behaviour may not reflect how people act in real life, giving the study low ecological validity [2].

9. A pilot study tests the research method on a small scale before the main study is carried out [1]. It allows the sociologist to identify and fix problems with wording, timing or procedure — such as a confusing question or a questionnaire that takes too long — before committing the time and cost of the full study [2].

---

Where marks are usually lost

• Confusing reliability with validity.
• Giving strengths of observation generally rather than of participant observation.
• Treating official statistics as objective fact.
• Suggesting random sampling for a group with no sampling frame.