1.20 Year 2 Research Methods - Probability and Significance Flashcards
Experimental Hypothesis
Non-Direction or Directions, one or the other is accepted
Null Hypothesis
State there will be no change in the conditions
Probability
Likelihood a certain event will occur, 0 means statistically impossible whereas 1 in statistically certain. This is written as a percentage represented as a decimal
Significance
How certain we are a difference or correlations exists, if a result is significant then the researcher can reject the Null and accept the experimental
Accepted level of significance is Psychology
P≤0.05 meaning there is a less than 5% chance the findings are due to chance. : P≤0.01 used when there is a possible human cost e.g. in a drug trail
Calculated Value (CV)
Result that has been calculated from the stats test, which is compared to the critical value.
Critical Value
Allows us to work out whether result is significant or not, and which hypothesis to accept, found in CV table. If significant we accept experimental hypothesis and vice versa.
Primary Data
Data researcher has found themselves
Secondary Data
Data researcher has found from another source
Meta-Analysis
Uses secondary data across multiple studies and analyses
The CV must be EQUAL or LESS than the CV to be significant when using
- Sign Test
- Mann Whitney U
- Wilcoxon
The CV must be EQUAL to or MORE than the CV to be significant when using
- Chi Squared
- Pearson’s R
- Spearman’s Rho
- Unrelated T-Test
- Related T-Test
Remember: RULE OF ‘R’
: Stats test with letter R means the CV must be MORE than the CV
Mean (Central Tendency)
Add all values and divide by the number of values
Strengths of Mean
- Eliminates anomalies
Limitations of Mean
- Still impacted by anomalies
Median (Central Tendency)
All values lined up and middle value picked
Strengths of Median
- No anomalies
Limitations of median
- Unrepresentative
Mode (Central Tendency)
Value used the most
Strengths of Mode
- No anomalies
Limitations of Mode
- Unrepresentative
Range (Dispersion)
Difference between lowest and highest results
Strengths of range
None
Limitations of range
- Impacted by anomalies
Standard Deviation (Dispersion)
Smaller the numbers the less it differs from mean
Strengths of Standard Deviations
- Thorough Analysis
Limitations of standard deviation
None
Type 1 Error
Most common. When researcher accidentally rejects the wrong hypothesis. Researcher thought they found a significant difference or correlation when they hadn’t – known as ‘false positive’ or optimistic error. Likely to occur when researcher is being too lenient
Type 2 Error
When a researcher accepts the null when it should have been rejected. Researcher thinks they didn’t find a significant difference when they did – known as a false negative or pessimistic error. Likely to occur when researcher is too strict
