Stats Recap

What do we mean when we say that results are 'significant'?

• Results are culturally important

• We have a large number of results to analyse

• Results are happening for a reason and not due to chance (a fluke)

• The results have revealed something that we didn't know previously

How do we check if results are signifcant?

• Rely on personal intuition
• Use anecdotal evidence to determine significance
• Conduct statistical tests

• Ask a friend for their opinion

What's the minimal level of signifcance that we use in Psychology?

(The probability of results being due to chance is less than....)

• 0.10
• 0.50
• 0.01
• 0.05

Sort these data sources into the level of data

Groups:

(a) Nominal

,

(b) Ordinal

,

(c) Interval

• Position of finish in a race

• Types of pets owned by students

• Length of a marathon in kilometers

• Level of agreement on a scale of strongly disagree to strongly agree

• Years of experience in a job

• Number of white shoes in a class

• Color of cars in a parking lot

• Rating of happiness on a scale of 1 - 10

• Names of countries in Asia

• Temperature in degrees Celsius

• Ranking of favorite movies from 1 to 10

• Time in hours and minutes

• Count of people who drive into college

• Weight in kilograms

• Answers to a closed ended question about favourite food

Complete the following saying "Chi is ​ (a)   and so is Rho, all the rest are ​ (b)   ". (This phrase should help you know when calculated values should be higher or lower than critical values.

• high
• low
• fine
• numbers
• fly
• flow

At the end of each stats test you end up with a CALCULATED VALUE that you compare against a CRITCAL VALUE. Which stats would need the calculated value to be higher than, lower than or equal to

Groups:

(a) Higher than the critical value

,

(b) Lower than the critical value

,

(c) Equal to

• Mann Whitney U

• Binomial Sign Test

• Chi Squared

• All of them!

• Wilcoxon Signed Ranks

• Spearman's Rho

Match the following

• Separate individuals in different groups

• Repeated Measures

• Similar people used in different groups

• Independent Measures

• Same participants across all conditions

• Matched Groups

Match the following:

• Independent Measures + Nominal Data

  (IN?)

• Mann Whitney U

• Independent Measures + Ordinal Data (IO?

• Chi Squared

• Repeated Measures + Ordinal Data (RO?)

• Wilcoxon Signed Ranks

• Correlation + Ordinal data (CO?)

• Spearman's Rho

• Repeated Measures + Nominal Data (RN?)

• Binominal Sign Test

Put the following steps for calculating a Binominal Sign Test in order:

• First of all, ignore any scores that stay the same

• Mark a + for any scores that have a positive change and a - for any scores with a negative change

• Count the number of +'s and -'s and identify the calculated value as whichever number is lower

• Compare the calculated value against the critical value in the critcal value table

• If the calculated value is lower (or equal to) than critical value then it is significant

Match the equation to the stats test

• Spearman's Rho

• 
• Chi Squared

• 
• Mann Whitney U

•