Inferential Statistics (research Methods) Flashcards
What are inferential statistics
Tests of statistical analysis
What are skewed distribution curves
Skewed distribution curves are when it is not symmetrical at the meab( or median or more) point. A skew can be positive and negative
What is positively skewed distribution
The long tail is going in a positive direction.
Low measure of central tendency
What is negatively skewed distribution
The long tail is going in a negative direction
High measure of central tendency
Define probability
The extent to which something is likely to happen or be the case
What are the different ways of expressing probability
Standard form
Decimal
Percentages
In psychology, how confident do you need to be that the results are significant and not due to chance
Atleast 95% confident Expressed like this P<0.05 P stands for probability < less than 0.05 is 5% expressed as a decimal
What are the two tests used in psychological research ?
Parametric tests
Non- parametric tests
What are parametric tests
Parametric tests can only be used for normally distributed data
What are non parametric tests
Non parametric tests are used for normal or skewed distribution.
Most research produced skewed data so non- parametric tests are more common
What are the parametric tests criteria
1) normally distributed data
2) variance is equal
3) interval data
4) no extreme scores
What are the non- parametric tests criteria ?
- when assumptions if the parametric test ms cannot be fulfilled
- when distributions are not normal
What are the 5 non- parametric tests
- Mann- Whitney U test
- Wilcoxon signed Ranks data
- Chi square
- Binominal sign data
- spearman’s Rho
What is the type 1 error
When the bill hypothesis is rejected and the alternate is wrongly accepted
What is type 2 error
When the alternate is wrongly rejected and the null is wrongly accepted
What are the Mann Whitney U test calculation steps
Research method: EXPERIMENT- Independent measures
Level of data: ORDINAL OR INTERVAL
Step 1 : rank all scores for both conditions as one whole group
Step 2: if more than one ppts has the same score add together their rank position and divide by the number of ppts
Step 3: add up the ranks for each group separately
Step 4 : using the formula for the smallest total value, work out the observed value
Step 5: compare to critical value
Wilcoxon signed rank test calculation steps
Level of data: ORDINAL OR INTERVAL
Research method: EXPERIMENT- REPEATED MEASURES
Step 1: find the difference between the scores in each condition
Step 2: rank the differences from lowest to highest
Step 3: add up how many positive and negative numbers there are in the differences column
Step 4: fund the observed value by using the least frequent sign and adding up the ranked differences
Step 5: compare to critical value table using number of ppts