Research Methods - Biological Psychology Flashcards
What can you do with histograms?
Display your Frequencies
What are the rules with histograms?
X Axis must be labelled frequency
Bars are coded so we know the group of tallies they stand for
Bars to touch
Blank space where frequency = 0
What are descriptive statistics?
Analyse findings from a sample
What are inferential statistics?
How sample’s resuults relate back to the target population where the sample was drawn
Why are inferential statistics needed?
Are essential to see whether the results support the null hypothesis or rejecting it in favour of the alternative hypothesis
Define the Mann-Whitney U-Test
For ordinal level data / experiments with independent groups design
Define the Wilcoxon Test
For ordinal level data / experiments with repeated measures or matched pairs designss
Define the Spearman’s Rho
For ordinal level data / Used for correlations
Define Chi-Squared
For Nominal level data / Analysing independent variables in categories (i.e. most observations)
How do you decide which test to use?
Step 1) Figure out the level of data collected (if nominal then, use Chi Squared)
Step 2)Is the study a correlation or a test of difference?(if correlation, then, use Spearmann’s Rho)
Step 3)Is the data independent or related?
(if independent groups design, then Mann Whitney - if otherwise, then Wilcoxon)
How do you know is the data is nominal?
Data is nominal if researchers identify categories of response and simply count how many times they occur
How do you know if the data is ordinal?
Data is ordinal if responses can be ordered from highest to lowest
What the four levels of data collection?
Ratio level, interval level, ordinal level and nominal level data
What is ratio level data?
A score on a scale.
The most important part of the data is that there is a meaning score of 0 (no data) which gives it a starting point.
What is interval level data?
Also a score on a scale, but can go into negative figures (is used for things like temperature)
Has no fixed stating point
What is ordinal level data?
A number score representing rank position (i.e. league tables) and can be turned into interval/ratio data by putting scores into rank order)
It demolishes distinctions between rankings
More psychological measures gather interval/ratio level data,so ranking is a must.
What is nominal level data?
Puts participants into categories (categorical data i.e. tally marks) to produce frequencies and is easily turned to percentages
Interval/ratio level data can be turned into nominal by putting data into a frequency table in order to turn it into a histogram.
What is the basic procedure for all inferential tests?
1) Calculate your observed value: meant to be as small as possible but in the case of Spearman’s Rho, it is the opposite (big as possible)
2) Choose your probability level: As these tests work out the likely-hood of your results as normality of results = acceptance of null hypothesis or very unlikely = refuting null for alternative hypothesis as shows a pattern at work
3) Find your critical value (part 1):Once selected a value for P and calculated your observed value, can consider a critical value table (different table for directional and non-directional hypothesis)
4)Find your Value of n or df: these are scores that represent the size of your sample (n = numbers of participants and df = degrees of freedom)
The more df, the more categories
5)Find your critical values (part 2): Compare observed values with critical value (found when counting along the table) to see if statistically significant
(Mann Whitney & Wilcoxon - U =or < critical value then is statistically significant
Chi Squared & Spearman Rho- id r =or > critical value then is statistically significant)
How unlikely do results have to be before you take them seriously and treat them as a pattern?
Decision summed up by P (probability)
P
