Psychology 203 Social Psychology

1
Psychology 203Social Psychology
2
A definition of social psychology

• "A scientific field that seeks to understand the
  nature and causes of individual behaviour in the
  social situation (Baron Byrne,1986).

3
The how to of psychology

• The task
• Provide you with the tools to describe,
  understand and critique the typical research used
  in this field.
• Provide some basic tools for understanding
  research
• Examine types of research and the problems they
  encounter
• Examine the strengths and weaknesses of the
  typical designs
• Examine the strengths and weaknesses of the
  specific experimental designed used in 203.

4

• What we want to establish with social
  psychological research is the cause(s) of a
  particular behaviour(s) in the social situation.
  
• Putting this is experimental terms what we want
  to know is whether it was the independent
  variable that brought about the change in the
  dependent variable,
• and, is this generalizable across situations and
  populations?

5

• Independent variableIn an experiment, the
  treatment or condition manipulated by the
  experimenter.
•  
• Dependent variableIn an experiment, any aspect
  of a subject's behaviour that is measured after
  the administration of a treatment the expected
  effect of a treatment.
•  

6

• Internal validity
• Asks the question Did the independent variable
  bring about the change in the dependent variable?
  The major problem here is when variables are
  confounded.
• Confound
• Variables are confounded when two or more
  variables are manipulated (usually the IV and
  some other variable) and we are unable to
  distinguish the effect of each on the DV.

7

• External validity
• Asks the question Can the results be
  generalised to other populations, times and
  settings?
• Often, but not always, we want to know that
  whatever we find will hold true for most, if not
  all people.

8

• Experimental realismThe extent to which the
  experimental procedures have an impact on the
  participants
• the extent to which events in the experimental
  setting are credible, involving and taken
  seriously

9
Experimental realism Helping Darley Latané,
1968

• Participants placed in cubicles to do study on
  communication
• Topic is personal problems associated with uni
  life.
• Take turns talking to others participants via
  intercom to avoid embarrassment. Es also leave
  room.
• First turn, one participant mentions that they
  have epileptic seizures.
• Second turn they start to have an audible
  epileptic seizure which lasts 3 minutes (180
  secs).
• Do you help? It depends on those around us.

10
Experimental realism Helping Darley Latané,
1968
85
52
62
93
31
166
11

• Mundane realismThe extent to which the
  experimental events in a controlled setting are
  similar to events in the real world.

12
Mundane realism Helping Darley Bateson (1973)

• Participants were young priests in training
• All asked to return later to give a talk on
• The parable of the good Samaritan
• The work issues facing young priests
• When they returned they were told that they were
• Early (If you have to wait over there, it
  shouldn't be for long)
• Very late (Oh, you're late they were expecting
  you a few minutes ago)
• On time (The assistant is ready for you, so
  please go right over )

13
Mundane realism Helping Darley Bateson (1973)

• On the way to give their talk all participants
  passed a person who was slumped in a doorway -
  unknown to them he was one of the experimenters.
  
• He looked shabby and was unconscious.
• Do they stop to help?

14
Mundane realism Helping Darley Bateson (1973)
33
42
0
80
50
25
15
True Experiments - Characteristics

• True experiments are characterized by the
  presence of the following
• A manipulation
• A high degree of control
• An appropriate comparison (the major goal of
  exerting control)
• Manipulation in the presence of control gives you
  an appropriate comparison.

16
True Experimental Designs

• Two characteristics of a True Experiment
• Random assignment to conditions
• The Great Equalizer Each person has an equal
  chance of being in any condition so that
  pre-existing differences among participants
  cancel each other out.
• Only the Independent Variable is varied
• Everything else is held constant across
  conditions thus any differences across
  conditions is due to the IV.

17
Threats to validity

• History occurrence of an event other than the
  treatment. The longer the study goes for the
  more likely this is to happen.
• Maturation participants always change as a
  function of time. Is change in behaviour due to
  something else? This can be biological or
  psychological.

18
Threats to validity

• Testing improvement due to practice on a test
  (familiarity with procedure, testers
  expectations)
• Instrumentation Any change in the observational
  technique that might account for the observed
  difference. Especially if humans are used to
  assess behaviour.

19
Threats to validity controlled by true
experiments (Campbell Stanley, 1966)

• Regression when first observation is extreme,
  next one is likely to be closer to the mean.
• Selection if differences between groups exist
  from the outset of a study.

20
Threats to validity controlled by true
experiments (Campbell Stanley, 1966)

• Mortality if exit from a study is not random,
  groups may end up very different. That is, the
  people who drop out are different to those that
  stay in.
• Interactions with selection
• Maturation
• Instrumentation (ceiling effects)

21
Threats to validity not controlled by experiments

• Contamination
• communication of information about the experiment
  between groups of subjects
• Diffusion of treatment
• control subjects use information given to others
  to change their own behaviour.

22
Threats to external validity

• Threats to external validity
• best way to deal with this is replication
• Hawthorne effect
• changes in a persons behaviour brought about by
  someone significant showing interest in them.
• a special kind of reactivity.
• name stems from studies of productivity at
  Western Electric Company, Hawthorne, Illinois,
  1924-1932.

23
Hawthorne study results

• Field study looking (in part) at illumination and
  performance
• Two groups one has lighting increased the other
  is left in the dark
• Both groups show increased productivity.
• Exp is restarted and this happens again.
• Despite a 70 reduction in lighting, the exp
  groups continues to show improvements in
  performance.

24
Threats to external validity

• Rosenthal effect/self-fulfilling prophecy
• Experimenters may inadvertently influence
  participants to do better or worse.
• This can be
• Active verbal or non-verbal behaviour
• Passive appearance, age, sex, race, dress.

25
Threats to external validity

• Rosenthal and Fode, 1963
• 12 experimenters were each given five rats who
  were to be taught to run a maze with the aid of
  visual cues.
• ½ were told their rats had been specially bred
  for maze brightness
• ½ were told their rats had been specially bred
  for maze dullness.
• There were no differences between the rats
  assigned to each of the two groups.
• Rats who had been run by experimenters expecting
  brighter behaviour showed significantly superior
  learning compared with rats run by experimenters
  expecting dull behaviour

26
Threats to external validity

• Selection x Treatment interaction.
• Does the effect of your treatment interact with
  characteristics of the experimental participants?
• That is, do your results generalize from the
  type of subjects you used in your research to
  other types of subjects?

27
Summarise the threats to IV

• M ortality
• R egression
• M aturation
• S election
• H istory
• I nstrumentation
• T esting

28
Understanding designs

• To make things easier, the following will act as
  representations within particular designs
• X--Treatment
• O--Observation or measurement
• R--Random assignment

29
Pseudoexperimental design

• The One Shot Case Study
• This is a single group studied only once. A group
  is introduced to a treatment or condition and
  then observed for changes which are attributed to
  the treatment
• X O

30
Pseudoexperimental design

• A total lack of control. We have no idea of
  causality.
• The error of misplaced precision where the
  researcher engages in tedious collection of
  specific detail, careful observation, testing and
  etc., and misinterprets this as obtaining good
  research.
• History, maturation, selection, mortality and
  interaction of selection and the experimental
  variable are all threats to the internal validity
  of this design

31
Pseudoexperimental design

• This is a presentation of a pretest, followed by
  a treatment, and then a posttest where the
  difference between O1 and O2 is explained by X
• O1 X O2

32
Pseudoexperimental design

• History
• between O1 and O2 many events may have occurred
  apart from X to produce the differences in
  outcomes.
• Maturation
• between O1 and O2 students may have grown older
  or internal states may have changed.
• Testing
• the effect of giving the pretest itself may
  effect the outcomes of the second test

33
Pseudoexperimental design

• Instrumentation
• The people or circumstances in which the testing
  is done may produce the changes.
• Regression.
• The change from time one may be due to the nature
  of scores people were selected on.

34

• There are several common designs used in social
  psychology
• Correlational research
• Field studies
• Field experiments
• Laboratory experiments (between and within
  subject designs).

35
Correlational  

• Gathers information about two or more variables
  without researcher intervention. For social
  psychologists this usually involves surveys or
  questionnaires.
• Strengths
• Estimates the strength and direction of
  relationships in the natural environment

36

• Weaknesses  
• Does not permit determination of cause-and-effect
  relationships among variables

37
Field studies (or quasi experiments)

• Gathers information about the behaviour of
  people who experience some real-world (natural)
  manipulation of variables in their environment.
  There is no matched or controlled assignment -
  you take the groups as you get them.

38

• Strengths
• Permits the study of real-word (full strength)
  variables. Hence it has good mundane realism and
  good external validity.
• Can easily be used to study practical issues

39

• Weaknesses
• Often difficult to obtain cause-effect
  relationships because of the inability to control
  environmental variables
• You may have to measure many different variables
  to eliminate these possibilities.

40
Quasi experimental design

• Common design in education because usually cant
  randomize assignment of students to classes
• pre-test measures whether initial groups are
  similar on tested variable could also match
  subjects based on pre-test but may miss other
  factors which could impact results (e.g., better
  teacher in one group)

41
Threat to IV

• Factors controlled by inclusion of a control
  group
• history
• maturation
• testing
• instrumentation (assuming same for both groups)
• statistical regression

42
Threats to IV

• Factors still NOT controlled
• Selection
• Mortality (if any)
• Selection x Treatment interaction

43
True Experimental Designs

• The Pretest-Posttest Control Group Design
• This designs takes on this form

44

• The Posttest-Only Control Group Design
• This design is as

45
Field experiments

• The IV is manipulated and its impact on the DV
  is measured in a natural setting. This is usually
  achieved by the use of control groups, and
  matched or random assignment of subjects to the
  control and experimental group.

46

• Strengths
• Permits the determination of cause-effect
  relationships (i.e., internal validity is good).
• The natural setting usually means high mundane
  realism and often, therefore, good external
  validity.

47

• Weaknesses
• Practical difficulties are the most common, i.e.,
  it is harder to control all the variables in the
  natural environment.
• Often it raises ethical issues such as invasion
  of privacy.

48
Laboratory experiment

• This is research in which the variance of all or
  nearly all of the possible influential variables
  not pertinent to the immediate problem of the
  investigation is kept at a minimum.
• This is done by isolating the research in a
  physical situation apart from the routine of
  ordinary life and by manipulating one or more IVs
  under controlled conditions and assessing their
  impact on one or more DVs.

49

• Strengths
• Permits determination of cause-effect
  relationships (i.e., it has good internal
  validity assuming no drop-out from the two
  groups).

50
Threats to IV

• All but mortality accounted for.

51

• Weaknesses
• Almost all concern External validity
• Because the data is obtained in an artificial
  environment it may lack generalizability to the
  real-world where lots of variables impinge upon
  us at the same time.

52

• Limited to certain groups who can be brought into
  the laboratory situation.
• The experimental situation may create a situation
  where subjects behave differently than they would
  in the real-world (e.g., to please the
  experimenters).
•  

53
Threats to EV

• Of course, none of this rigor tells us whether
  this effect may also occur in the outside world
  if that's your aim.

54
To help provide external validity

• Random samplingSubjects are selected at random
  from the population (e.g., getting a random
  number generator to produce random phone
  numbers).
•  
• Stratified samplingSelecting a proportional
  sample from each different stratification in
  society (e.g., selecting 50 males and 50
  females 10 low income earners and 30 low
  income earners).

55
Between subject designs

• There are two (or more) values of the IV and
  only one value is administered to each group.
• One group receives one level of the independent
  variable and the other group either receives
  another level of the same IV or nothing (as in a
  control group).
• We assume that the other variables, which could
  affect the DV, are distributed equally between
  the group.

56

• You can help provide internal validity by
  ensuring that the variables that are not
  controlled are distributed equally between the
  groups. To ensure this we can use
• Random assignment where, for example, a coin is
  tossed for each subject and if it is a 'head'
  they go to group A.
• Matched assignment where two subjects matched on
  as many important variables as is possible are
  assigned one to each group (e.g., one male,
  middle-aged, high income earner goes to each
  group).

57
Within subject design

• Two (or more) values of the IV are administered,
  in turn, to the same subjects. Each person
  becomes, in effect, their own control.

58

• Strengths  
• Less people needed as they are there own
  controls.  
• Cause and effect relationships are often easier
  to find, even when the effect of the IV is small,
  because error variance (within the groups) is
  kept constant.  

59

• Weaknesses
• Practice effects can be confused with the effects
  of the IV

60

• Carry over effects arise when receiving one value
  of the IV influences reactions to the second
  value of the IV

61

• Sensitisation effects arise when, after receiving
  one value of the IV, subject 'work-out' (or think
  they have) the purpose of the experiment. Then
  they modify their responses accordingly.

62
(No Transcript)
63
The Decision Criterion
Critical Region
64
Compare Mean to Null Hypothesis
Compute z-score of where sample mean is located
relative to the hypothesised population mean
sM SEM is defined by
65
Compare Mean to Null Hypothesis
sM SEM is defined by
66
Observed differences

• When does a difference make a difference?
• Are all differences the same?
• Two sets of means can be different by the same
  magnitude but they can be treated very
  differently depending on the sample sizes.
• Sometimes deciding whether a difference is
  important needs to be assessed with respect to
  previous research and theory.

67
What is the Standard error?

• Standard
• Error

68
The distribution of sample means for random
samples of size (a) n 1, (b) n 4, and (c) n
100 obtained from a normal population with µ 80
and s 20. Notice that the size of the standard
error decreases as the sample size increases.
69
Standard Error of the Mean

• The SEM is affected by sample size as the formula
  suggests.
• Consequently as the sample gets larger our
  estimate of the true mean gets better and we can
  be more confident about where the true mean lies.

70
Sampling Distribution of the Mean (SEM)Summary

• Repeated taking samples from a population
  produces a distribution of sample means which is
  bell shaped.
• The properties of the distribution are that most
  samples yield means close to the true population
  mean but they vary.
• Plotting the sample means repetitiously would
  produce a sampling distribution of means that is
  normal.
• Normal distributions have certain properties that
  are useful.
• For one thing we can specify a criterion defining
  oddness or, if you like, difference.

71
Standard Error of the Meanof the difference
between samples

• We can apply the same logic to the distribution
  of differences between means that we get when we
  sample two means from a population.
• Sometimes the means will differ by chance but
  how likely is the difference to be due to chance?
• Finally, remember that were looking to test a
  null hypothesis

72
t-test

• t-test is used to test hypothesis about µ when
  the value for s is unknown
• The formula for the t statistic is similar in
  structure to the z-score, except that t statistic
  uses estimated standard error from the sample,
  not population standard error

73
The t Distribution
We use t when the population variance is unknown
(the usual case) and sample size is small (Nlt100,
the usual case). If you use a stat package for
testing hypotheses about means, you will use t.
The t distribution is a short, fat relative of
the normal. The shape of t depends on its df. As
N becomes infinitely large, t becomes normal.
74
(No Transcript)
75
T-Test for Dependent Samples
Our hypotheses Ho ?D 0 HA ?D ? 0
To test the null hypothesis, well again compute
a t statistic and look it up in the t table.
76
Steps for Calculating a Test Statistic
77
t-test Formula - I

• Standard Error of X is estimated from the sample.

Standard Error of X is calculated using the
sample standard deviation.
78
t-test Formula - II
79
Related Samples t-test

• Scenario 1 I want to test the effectiveness of
  a new relaxation technique.
• 25 people take a stress test then go through my
  relaxation techniques then take the stress test
  again.
• Comparing same people at 2 points in time.

80
An example

• Take a random sample of 10 students from the
  class and compare their scores for the two tests.
• Any improvement?

81
Computing the t statistic for related (paired)
samples

• Based on difference scores rather than raw
  scores.
• Difference Score
• D (X2-X1)
• Must find for every subject.
• Then compute D-bar (mean of differences)

82
t-statistic for related samples

• t-statistic
• Standard
• error of D is
• df is calculated the same n - 1.

83
Hypothesis Testing Steps

• Step 1 Formulate the hypotheses
• Step 2 Set the decision region
• df n 1
• Step 3 Calculate the t-statistic
• Step 4 Make and report your decision.

84

• Mean difference 5.6
• Standard deviation 5.3996
• Standard error of the mean 5.3996/sqrt of n (10)
  
• SE 5.3996 / 3.16
• SE 1.7075
• t 5.6/ 1.7075
• t 3.28

85
(No Transcript)