questionaires Flashcards
what are investigator effects (experimenter bias)
Anything that the investigator does which has an effect on the PPs performance in a study other than what was intended.
Conscious or unconscious e.g. A really good looking experimenter.
We can try to eliminate investigator effects by implementing a double blind procedure.
what is a double blind procedure
neither the pps or the experiment knows the precise aims of the study
-this reduces experimenter effects as it prevents investigators from giving pps clues (consciously or subconsciously) as to which condition they are in
what is a single blind procedure
where the pps are not informed of the condition that they are in
-this helps to eliminate demand characteristics
how can investigator effects
Physical characteristics of investigator – e.g. age or ethnicity. Male participants may be unwilling to admit sexist views in front of a female researcher.
Less obvious personal characteristics. For example, participants may respond differently to someone with a stern voice.
Investigators may be unconsciously biased in their interpretation of the data and find what they expect to find.
characteristics of a questionnaire
consist of closed/fixed choice items
closed/fixed choice - easier to score/quantify (quantitative data) but restricts pps answers
-open ended questions - more realistic as more scope to answer questions in our own way - more detailed
likert scale - 5-7 point scale
easy to analyse, quicker, easier
what are the 4 reasons foudn by coolican as to why it is hard to produce a good questionnaire
complexity - item difficult to understand
ambiguity - item can be interpreted in more than one way
double barrelled items - items can contain 2 questions and ask for y/n response
leading questions - items that can contain within its implications that a certain response is expected
what should be considered fro constructing a questionnaire
Aim: An exact aim helps then it’s easier to write questions to address this.
Length: short and to the point to reduce drop out rate
Previous questionnaire – use as a basis if they’ve been successful
Question formation – should be concise, unambiguous and easily understood
grammar + wording need to be accessible
can question be interpreted in more than one way
Pilot study
Measurement scales – statements on which participants rate levels of agreement or disagreement
how can we decide how good any questionnaire is?
-standardisation - tests that have been administered to a large representative sample so the individual scores can be compared against others
doesn’t have to be representative of general population but of target.
-reliability - extent to which questionnaire provides consistent findings
-test-retest technique - a number of individuals are given the same questionnaire on two different occasions and the scores are correlated
high correlation = reliability
-split half technique - scores from one half of test compared with other
high correlation = reliability
-validity - questionnaire is measuring what it claims to measure
e.g. questionnaire assessing aggression in children, valid if those who score higher are rated more aggressive by teachers
strengths of questionnaires
-large amounts of data collected and analysed (at relatively little cost) so conclusions can be drawn
in little time
-explore most aspects of personality, attitudes, and beliefs, making questionnaires very versatile
-many questionnaires have high reliability and reasonable consistency
-lack of investigator effects - can be completed without researcher present
weaknesses of questionnaires
- social desirability bias
- many questionnaires poorly constructed so lack reliability and validity
- often used with correlation studies so can’t establish c+e
- pps need high literacy skills to understand meaning of items
- biased sample - certain types of people more willing to fill in questionnaires
- some questionnaires may require detailed answers and closed questions may be given
