Investigation 2 Flashcards
What is the title of investigation 2
A correlation study involving the stroop test
What is a correlation study
Used to analyse the relationships or association between 2 continuous variables ( co-variables)
Scatter diagram can be used to illustrate a correlation
The closer the coefficient is to 0 the weaker the coefficient - the closer it is 1 (+1/-1) the coefficient is strong
What is the stroop test
The stroop test effect is the delay in reaction time between automatic and controlled processing of information, in which the names of words interfere with the ability to name the colour of ink used top print the words (tam 2015)
What are the operationalised co variables
1- creativity scores on the Kellogg creativity questionnaire out of 116
2- time taken in seconds to accurately complete the stroop test
What are the confounding variables
Dyslexia
Caffeine intake
Sleep
Food intake
Colour blindness
How to overcome confounding variables
Discount certain scores
What are the extraneous variables
Time of day
Weather
Noise levels
How to overcome extraneous variables
Do it in a quiet classroom
Test retest
How to overcome extraneous variables
Do it in a quiet classroom
Test retest
Which hypothesis do we use
Alternative hypothesis - there will be a negative correlation between creativity scores out of 116 on Kellogg creativity questionnaire and the time taken to accurately complete the stroop test in seconds
What’s the null hypothesis
There will be no correlation between creativity scores out of 116 on Kelloggs creativity questionnaire and the time taken in seconds to accurately complete the stroop test
Methodology
Correlation
Why have we used a correlation
Because we are testing for a relationship not a difference
+/- of a correlation
+easily replicated, no manipulation, quicker
- doesn’t establish cause and effect
What’s the target population
P’s over 11 years old
What’s the sampling frame
Students over 11 at Norte dame / teachers
What sampling technique did we use
Opportunity sampling
Justify the sampling technique chose
Quicker and easier then stratified as we don’t need a list of the sampling frame
How to ensure privacy
Results wont be available to everyone
Don’t ask personal info - stick to questionnaire
How to ensure confidentiality
Code names e.g house number/bday
How to reduce chance of deception
Will reveal the aims in brief and debrief
How to ensure right to withdraw
Clearly state the right to withdraw at the start and throughout
Allow a 48h reflection period
How to ensure informed consent
Some p’s are under 16 = no valid consent
Needs headteachers consent on their behalf ( consent by proxy)
How to reduce protection from harm
Offer websites in the debrief for support or teachers emails as they may feel stressed or embarrassed about a low score
Stroop test may be stressful
How to ensure internal reliability is high
Standardise procedures
Same questionnaire, stroop test, researcher, wording of questions
Could check for external reliability
Test retest after 6 months (same p’s and same method) then compare
Validity issues
Social desirability - as its a questionnaire they may lie to seem more creative
Demand characteristics- we give a full brief/debrief revealing the aims so they may guess the aims
How to overcome the validity issues
SD- anonymous questionnaires so p’s cant be identified = less likely to lie
DC-tell them the aims but not the hypothesis
What measures of central tendency can we use
All - doesn’t matter if there’s a nonsense value at its time in seconds and all of the central tendency are quite equal
Which measures of dispersion can we use
Both as there is no huge anomalies
What graphical representation do we use and why
Scatter diagram as its a correlation
What 3 steps determine which inferential statistic we use
-correlation
-ordinal and above
-relationship
What 3 steps determine which inferential statistic we use
-correlation
-ordinal and above
-relationship
Which inferential statistic did we use
Spearman’s rho
What was the observed value and what did it show
-0.39 shows a weak correlation coefficient
Were results significant
Non significant because the observed value is not equal to or greater than the critical
We should accept the null and reject the alternative - 0.05% due to chance
Conclusions
Findings are non significant and we should accept the null
Results show a weak correlation and any relationship is due to chance
Ways to improve
Test retest after 6 months to secure reliability
Use a bigger phone
Use more ps from varied age group for a better sample
Use a better environment like a quiet classroom to control ev’s like noise
Ways to improve
Test retest after 6 months to secure reliability
Use a bigger phone
Use more ps from varied age group for a better sample
Use a better environment like a quiet classroom to control ev’s like noise
