Self report methods are ways to obtain data by asking participants to provide information about themselves. There are two types of self report methods:
Questionnaires are self report methods that use written questions.
Interviews are self report methods in which participants verbally reply to questions asked by an interviewer.
A rating scale is a simple numerical scale on which a participants can indicate a choice by selecting a number. eg: 'on a scale of 1-10...'
Likert scales begin with a statement and ask participants to respond by choosing an option typically from 'strongly disagree, disagree, don't know, agree, strongly disagree.'
A Pilot study, which is a small-scale trial run of a method, can be carried out to identify any practical or ethical problems and resolve them.
Strengths for rating/likert scales:
- easy to respond to so data can be collected quickly, making data more reliable and easier to conduct on a wide sample making the data more generalisable.
- they produce quantitative data which is easy to compare and analyse.
- they can be tested for reliability (by test-retest or split half techniques) and improved by changing and removing unreliable questions easily.
Weaknesses of rating/likert scales:
- only quantitative data produced which may lack detail and participants are not able to express opinions fully, lowering validity.
- response bias (participants maybe consistantly giving answers in the middle of the scale or at one end)
- the points on the scale are only relative (ordinal data) i.e the gaps between the points are not equal so the data should not be used to calculate a mean.
- rating scales cannot be used to measure complex variables that require more than a simple numerical response such as attitudes.
- the middle value in a likert scale is ambiguous.
Closed questions offer a small number of explicitly stated alternative responses and no opportunity for participants to expand of the answers. produce quantitative data.
STRENGTHS: easy for participants to respond to, so large amounts of data can be collected quickly making data more reliable and easier to assess a large sample which can make the research more generalisable. the quantitative data produced is easy to compare and analyse. (modes, medians can be found and put on graphs etc)
WEAKNESSES: only quantative data can lack detail thus lowering validity. Risk of response biases (such as constatly saying 'yes').
Open questions allow participants to give full and detailed answers in their own words, no categories or choices are given. Produces qualitative data.
STRENGTHS: Qualitative data provides detail so participants can fully express their answers, raising validity.
WEAKNESSES: Qualitative data can be time consuming to analyse as themes need to be identified and extracted. Interpretation of the data can be subjective, potentially reducing inter-rater reliability. Findings may be less generalisable.
A structured interview asks predominantly closed questions in a fixed order. The questions are likely to be scripted so they are standardised.
A semi-structured interview uses a fixed list of open and closed questions, although the interviewer can introduce additional questions if requred.
An unstructured interview generally begins with a standard question for all participants but from there on, questions depend on the respondents answers. There may be a list of topics for the interviewer to cover.
Test-retest is a measure of reliability that uses the same test twice. Split-half is a measure of reliabilty that compares two halves of a test. In both of these measures, If the participants' two sets of scores correlate well then the test has good reliability.
Correlation coefficient: expressed as an 'r' value between 0 and 1. A high value (eg. 0.8) means that there is a strong correlation, 0 would mean no correlation. A higher value suggests better reliability.