Lecture 11 15/05/13
Part A: Intelligence and measurement
Types of tests
Aptitude
> Measure what you may be able to do in the future
Achievement
> Measure what you can do now
Intelligence
> Measure general cognitive functioning
Personality
> Measure as
Lecture 3 13/03/13
Part A: Aggression and prosocial behaviour
Why are people so aggressive?
Biological explanations:
We know that the amygdala is particularly important in regulating
aggression
Alcohol reduces the ability of people to monitor the conseq
Part B: Organisational psychology
Definitions
Industrial and organisational psychology is a speciality area within the
broad field of psychology that studies human behaviour in work settings
Practice is based on systematic research evidence and, in theo
Risk: the risk that an individual within a group falls into an undesirable category
is simply the proportion in that category
risk = number in category
(expressed as %)
total number in group
Relative Risk: influence of an explanatory variable on a risk
re

Equation for a straight line y = b0 + b1x (where b0= slope, where b1 = intercept [y
when x = 0])
Want to model the mean (or expected value) of response y as a straight line function of
x
Estimated Regression Line
o y = b0 + b1x
o Can use to predict app
7. Experimenter effects mistakes in data collection
8. Ecological validity results in controlled lab may not match what we would observe in
the real world
Lecture 7
Trial
 Each repetition of an experiment
Outcomes
 Results we might see
Event
 Subse
OneSample TTest:
Interested in the population mean,
Want to test the hypothesis that the population mean is equal to some
particular value, 0
Null hypothesis  H0: = 0
Alternative hypotheses:
1. H1: 0 (twosided)
2. H1: < 0 (one sided)
3. H1: > 0 (
1. Positive as explanatory variable increases, as does response variable
2. Negative as explanatory variable increases, response decreases
3. *A strong relationship is one which does not have much variability*
Pearson Correlation Coefficient (r)
 Summari
The ChiSquare Test: formal statistical approach used to analyse categorical
variables
One categorical variable: GoodnessofFit Test
Two categorical variables: Contingency Table Analysis
GoodnessofFit Test
Expected counts: what to expect when the nul
Procedure:
1. Calculate the difference for each pair
2. Rank absolute differences
3. Calculate the sum of ranks
4. If sum is too small or too large (depending on the alternative hypothesis),
reject the null hypothesis
Note: if the sample size is large eno
Transformation
If the relationship is nonlinear, can possibly find some form of mathematical
transformation to make the relationship between the two variables linear, which
then enables the assumption of linearity
Transformation: taking some function of
Analysis of Variance (ANOVA): extension of the twosample ttest, used to
simultaneously compare more than 2 independent means
Start point: assume that these samples come from populations which have the
same mean (statement of no effect no difference in m
Written as Prob > F
Less than 0.0001, therefore we have evidence to reject H0 so we can
expect that, in the population, the mean change will not be the same
across all groups
Barletts Test
 Written as Prob > chi2
 0.373, therefore no evidence against