Week Six

Question 1

why use more than two groups?

Answer 1

• if there are more than 2 groups of interest + a control
• including 3 or more groups can allow de-confounding
• looking for nature of relationships

Question 2

relationships

Answer 2

• linear
• curved
• quadratic

Question 3

levels of IV

Answer 3

○ Determined by type of relationship
○ Linear needs at least 3 points
- How far apart should the levels be?
○ Proportionately across spectrum
○ Allows for clear examination of levels of IV
§ Only applies to IVs that are based on measurement, rather than categories.

Question 4

analysing multiple groups design

Answer 4

• ineffective to use t tests when we have more than 3 conditions.
• using multiple t tests would result in massive type 1 error.

Question 5

analysis of variance

Answer 5

ANOVA
- We can no longer look at the difference between means as there is more than 2 groups
F ratio is needed

Question 6

f ratio

Answer 6

• F ratio can also be used in 2 groups
  
  • The larger the f ratio the more likely it is to be significant.
  • Will be larger when the numerator is larger.

Question 7

ANOVA and variability

Answer 7

• A well designed experiment with a single IV there will only be 2 sources of variability in the data
  ○ Variability due to the effects of manipulating the IV
  ○ Variability due to sampling error.
  
  • ANOVA isolates these sources of variability to see is sampling error can account for any apparent differences in scores between groups.
  • ANOVA does this by looking at the ratio of the variability between groups compared with the variability within groups.

Question 8

between groups variability

Answer 8

• Need to calculate the variance between group means.
  
  • Calculate this variance by looking at how group means vary around the grand mean.
  • Find the variance of groups means around the grand mean.

Question 9

within groups variability

Answer 9

• Need to calculate variance within groups
  
  • Calculate variance for each group separately
    ○ Variance of individual scores around their group means.

Question 10

calculation of variability and ANOVA

Answer 10

• The WG variability is calculated from various between groups scores of participants treated alike
  
  • The BG variance is calculated from variations in the mean scores between levels of the IV.
  • The WG variability and BG variability are potentially due to difference caused
    ○ WG variability must be due to sampling error
    ○ BG variability may be due to both
    § The effect of the IV and sampling error.

Question 11

hypothesis testing and F

Answer 11

Ho= all groups means are the same and the IV has no effect. 
H1= at least one group mean is different and thus the IV has an effect.

Question 12

Ho true

Answer 12

if the null hypothesis is true then
BG vari= sampling error (E)
WG V = sampling error (E).
BG/WG=E/E=1

Question 13

H1 true

Answer 13

BG V= error + treatment effect
WG = error
BG/WG= E+ treatment/E= >1.

Question 14

means square

Answer 14

same as variance in SPSS.

Question 15

sample variance (descriptive)

Answer 15

= sum (X-M)^2/n

Question 16

population variance (inferential)

Answer 16

= sum (X-M)^2/n-1

Question 17

Df total

Answer 17

n-1

Question 18

df between

Answer 18

number of groups - 1

Question 19

df within

Answer 19

n- number of groups

Question 20

calculating variance and the f ratio

Answer 20

MS= ss/df 
f = MS between / MS within