What an interview is

1. An interview collects data by asking someone questions out loud, usually face to face but sometimes by phone or video call.
2. It is a self-report method, so the participant describes their own thoughts, feelings and behaviour.
3. The interviewer writes the answers down or records the audio, which needs the participant's consent first.
4. A recording is typed up afterwards, and that written transcript is what gets analysed.
5. Two types cover it: structured and unstructured.
6. How far the questions are fixed in advance is what decides which of the two types an interview is.

Structured interviews

1. A structured interview uses the same questions, in the same words and in the same order, for every participant.
2. The questions are written before the study starts and the interviewer reads them from the sheet without adding anything.
3. Most of the questions are closed or use a rating scale, so the answers arrive as numbers or as short set responses.
4. That makes the data mainly quantitative, ready to total, average and put in a bar chart.
5. It suits a question about how common something is, or about comparing two groups, such as whether Year 11 pupils revise more hours than Year 10 pupils.
6. Strength: the standardised script can be repeated word for word, which makes the findings testable for reliability.
7. Strength: a set script needs less training, so several interviewers can share the work and still collect comparable answers.
8. Weakness: fixed questions cannot follow up an unexpected answer, so anything the researcher did not think to ask goes unrecorded.
9. Weakness: forcing a complicated feeling into one of four options can misrepresent what the person means, which costs validity.

Unstructured interviews

1. An unstructured interview starts from a topic and an opening question, then follows wherever the answers lead.
2. Nothing after the first question is planned, because each question comes out of what the participant has just said.
3. The questions are open, so the data are qualitative and arrive as the participant's own words.
4. It suits a question about why people act as they do, or how something feels, such as what living with insomnia is like day to day.
5. It is also the method used inside a case study of one person, where depth of detail is what the study needs.
6. Strength: the interviewer can follow up anything that matters, which reaches reasons a fixed script would have missed and raises validity.
7. Strength: a relaxed conversation puts people at ease, so they say more about a sensitive topic.
8. Weakness: every interview runs differently, so none of them can be repeated exactly and reliability is low.
9. Weakness: turning pages of speech into findings involves judgement, so two researchers reading the same transcript can pull out different themes.
10. Weakness: it takes a trained interviewer and a long session, so only a few people can be reached.

Features shared by both types

1. In both types the interviewer can explain a question the participant has not understood.
2. Both are slow to run, so an interview study reaches few people and the sample is often too small to generalise from.
3. Because the topics can be personal, such as sleep, mood or school, consent and the right to withdraw are set out before the interview starts.
4. An interview has good ecological validity, because talking about your own life is an ordinary thing to do.

Bias in interviews

1. Social desirability bias affects interviews strongly, because saying "nine hours on my phone" out loud to a real person is harder than writing it on a form.
2. Interviewer bias is the interviewer's own tone, age, accent or expression nudging an answer one way.
3. Bias in questioning comes from a leading question, since "you do revise, don't you?" invites the answer it wants.
4. Promising confidentiality before the interview starts reduces social desirability bias, because a participant who trusts the promise answers honestly.

Turning interview answers into data

1. Closed and scaled answers from a structured interview become numbers that can be summarised straight away.
2. Worked example: five pupils report 4, 6, 6, 9 and 10 hours of revision in a week.
   1. The total is 35 hours, so the mean is 35 divided by 5, which is 7 hours.
   2. In order the middle value is 6, so the median is 6 hours.
   3. The value appearing most often is 6, so the mode is 6 hours.
   4. The range is 10 minus 4, which is 6 hours, and the range measures spread rather than an average.
3. Answers to a rating scale go into a frequency table and then a bar chart, with a title and both axes labelled.
4. Qualitative answers from an unstructured interview are handled by picking out themes and quoting the words that show each one.
5. Counting how many participants mentioned a theme turns qualitative material into a number that can be reported.

Choosing the type of interview

1. An aim that involves counting or comparing groups points to a structured interview.
2. An aim that involves exploring reasons or experiences points to an unstructured interview.
3. The type is stated plainly, such as a structured interview with the same six questions for every pupil.
4. One or two of the questions are written out in full, so the wording can be checked for leading language.

Turning this into marks

1. When a type and a reason are both asked for, keep them as two separate points, one naming the type and one naming the kind of data it produces.
2. Write out one question that would actually be asked, because "I would interview them" does not secure the knowledge mark.
3. Tie the weakness to the type chosen and to the topic, such as pupils understating late nights when asked out loud.
4. The mark most often lost is quoting a mean without its units, so write 7 hours rather than 7.