Questionnaires Flashcards
What is a rating scale?
Respondents asked to chose from a range of options, which are scaled between 2 extremes
What is a likert rating scale?
Respondents rating their opinion by circling a number to show the extent to which they agree or disagree. (even numbers force them to show degree of opinion)
Strongly agree (5) - (4) - Neutral (3) - (2) - Strongly disagree (1)
What is a semantic rating scale?
Respondents indicate where their attitude falls between extremes (2 bipolar adjectives)
Do you see yourself as strong or weak?
strong : : : : : : : : : : weak
(numbered 1-10 (unseen))
What is ranked data scale?
Respondents ranks their choices relative to other options.
Each preference should be given a weighting / score
First choice = highest rating etc.
What is the difference between qualitative and quantitative data?
Quantitative = numerical
Qualitative = non-numerical
What is the difference between open-ended and closed questions?
Open-ended questions = don’t involve pre-set answers, but allow respondent to answer freely, by elaborating and justifying answers. get attitudes, thoughts, ideas, experiences
Closed questions = questions that have pre-set fixed answers that a respondent selects by circling / ticking their opinion
An issue with questionnaire design is social desirability. What does this mean?
Respondent doesn’t give genuine answer, but one that depicts them in a more favourable light (one that agrees with prevailing social norms)
How can we avoid question construction being an issue of questionnaire design?
Questions should not be too technical, ambiguous or complex.
Questions should not mislead respondent to giving a particular answer.
Questions should not ask personal questions.
In likert scales, i.e. use even numbers so ppts. can’t chose middle option + can’t stay neutral.
How can we establish external reliability in questionnaires?
Test - retest:
Some people are given the same question are to complete again on a different occasion.
External reliability is established if the responses are very similar.
How can we establish internal reliability in questionnaires?
Split - half:
Split questions into 2 halves and compare + analyse findings from both halves.
If all questions measure the same concept, both halves should achieve the same score.
If not, suggests some questions may measure a different concept- NOT RELIABLE
What is face validity in questionnaires?
Reviewing each question and deciding if it makes sense in terms of what is being measured
And only measures what it is supposed to measure.
What is predictive validity in questionnaires?
Being able to accurately predict the same construct in the future
e.g. an intelligence test with predictive validity- high intelligence score should correlate with educational success
If a questionnaire is a valid measure of construct, it should have predictive validity.
What is concurrent validity in questionnaires?
Comparing one questionnaire to another test measuring the same concept (should run alongside each other and agree with one another)
What is the difference between external and internal reliability?
External = the consistency of a measure or finding over time
Internal = the consistency of a measure within itself
