Transcript Experiments & Observations

Content Page
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Observational Method
Experimental Method
Experimental Design
Aims & Hypothesis
Variables: IV, VD & EV/CV
Operationalisation
Sampling Techniques
Factors Associated with Research Design
Reliability
Validity
Ethical Issues
Data
Suggestions for future/Improvements
Writing Procedure
Core Studies
Loftus & Palmer
Maguire
Baron-Cohen
Piliavin
Savage-Rumbaugh
Reicher & Haslam
Samuel & Bryant
Rosenhan
Bandura
Thigpen & Cleckley
Freud
Griffiths
Dement & Kleitman
Milgram
Sperry
1. Observational
Method
G541 & G544
Section B
1. Briefly outline the observational method used in Psychology
(4 marks)
2. Describe 2 observational studies in Psychology (8 marks)
3. Discuss strengths and limitations of using the observational
method to investigate behaviour. Use examples of
psychological research to support your answer (12 marks)
4. Compare observations with self reports. Use examples of
psychological research to support your answer
Key features of the observational
method
• In an observation, data are collected by someone observing
(watching) participants and recording what pts do or say.
• Sometimes the observer is present and sometimes the
observer is hidden behind a one-way mirror.
• Other recording techniques can be used, including video
recordings and CCTV footage.
• Observations may be conducted on their own or they may be
conducted as part of an experiment.
NB: What identifies an investigation as an experiment is
whether it has an IV and DV rather ran where it is
conducted or the details of the procedure
Observation in Core Studies
Study
Loftus &
Palmer
Baron-Cohen
SavageRumbaugh
Bandura
Freud
Dement &
Kleitman
Sperry
Approach
Observational
Role of observation in
Study
Observation
Nature & Use
Everything is left as normal, all variables are free to vary.
Exam behaviour without the interfering in any way
Design: Overt or covert observations
Advantages
•People tend to behave naturally (High EV)
•Information gathered is rich and full
•Can be used where other methods are not possible,
where variables cannot be manipulated.
Disadvantages
•Experimenter has no control over the situation
•Participants can be aware of being watched and this
can affect behaviour
•Problems of reliability due to bias or imprecise
categorisation of behaviour
•Problems of validity due to observers’ or coders’
assumptions
•Replication difficult
Informed consent
Difficulty debriefing
Privacy
Ethical Issues
Different Types of Observation
Types of Observation
Controlled
Observations
Conducted in a laboratory or classroom as part
of an experimental procedure. Control of Pts and
asked to do set tasks e.g. Bandura
Natural
Observations
Participants
Observations
•No control over pts and what they do e.g.
Piliavin
•Type of natural observation, but observer is
playing an active role and fully participating
member of the group.
Overt & Covert •Griffiths
•Bandura
Structured &
Unstructured
•Unstructured >Milgram
•Time Sampling > Bandura e.g. 5 sec intervals
•Event Sampling > Bandura
Activity 1
Strengths & Weaknesses of Observational Method
Observation
Strength
Weaknesses
Activity 1
Strengths & Weaknesses of Observational Method
Observation
Strength
Weaknesses
•High EcV > naturalistic
•Can establish C&E
•Rich, qualitative data
•Quan & Quali
•Covert observations unethical
•Easy to run
•Can’t control all variables.
Little standardisations.
Difficult to replicate.
•Low in DC’s (covert)
•High DC’s (overt)
•
Improving Observations
Inter-observer reliability:
• 2 observers
• Should be agreement on data collected
Activity 2
Strengths & Weaknesses of Observational Types
Strength
Controlled
Natural
Participant
Unstructured
Structured
Time Sampling
Event Sampling
Overt Sampling
Covert Sampling
Weaknesses
2. Experimental
Method
Section B
1. Briefly outline the experimental method used in Psychology
(4 marks)
2. Describe 2 laboratory experiments in Psychology (8 marks)
3. Discuss strengths and limitations of using the experimental
method to investigate behaviour. Use examples of
psychological research to support your answer (12 marks)
4. Compare laboratory experiments with field experiments.
Use examples of psychological research to support your
answer
5. Discuss the extent to which psychology can be a science (8
marks)
Brainstorm
• Write down everything you can remember about the
experimental method
Brainstorm
• Write down everything you can remember about the
experimental method:
•
•
•
•
•
•
•
Types
Design
Hypothesis
Variables
Control
Ethical Issues
Data Collection
Research that uses the experimental
method…
Key features of the
experimental method
• Theory: The aim of an experimental is to test a hypothesis
(prediction) with the aim of disproving or supporting it.
• Test: In order to test the prediction, it has to be established
that one variable (thing) has a measurable effect on another
variable (thing).
• Control: The study must be conducted under controlled
conditions so that the researcher can identify that the effect
that has been found is due only to an identified variable and
not to other factors that were not tested.
• Replication: In order for support for a theory to be retested it
is vital that any experiment can be replicated (imitated with
the same results) by others. This means that the method must
be identified precisely and be standardised so that it can be
imitated.
Key steps in an experiment
1. The experimenter comes up with a hypothesis
2. The experimenter designs an experiment to test that
hypothesis
3. The experimenter manipulates one factor (the IV) that s/he
has identified in his hypothesis as being likely to cause a
particular effect.
4. The experimenter measures the effect of this manipulation
(DV).
5. Other variables are controlled.
6. The experimenter analyses the difference in the mean
results obtained in each condition.
7. If a significant difference is found between means, this
supports the alternative hypothesis. If no significant
difference is found, the null hypothesis is retained.
Strengths
-Test hypothesis by manipulation of IV
-Scientific: follows standardised
procedures > enables replication
-Control over extraneous variables
-Produces quantitative data which can
be statistical analyses to ensure
meaningful comparison.
Limitations
-Low ecological validity – removed
from real life
-Small sample reduces generalisability
-Causes stress/anxiety to participants
(ethical issues)
-Doesn't collect qualitative data
therefore reductionist.
Laboratory Experiment
Nature & Use
Advantages
Disadvantages
IV manipulated to observe the effect on
DV, under controlled conditions.
 Establishes causal relationships
 Allows for replication
 Good control over confounding
variables
 Artificiality: Mundane realism and
experimental realism
 The effects of being observed:
Demand characteristics,
(participants) evaluation
apprehension (researcher)
Field Experiment
Nature & Use
Advantages
Disadvantages
Investigate causal relationships in more
natural surrounding
•Establishes cause and effect
relationships
•Allows for replication
•Behaviour of participants more typical
than in a laboratory experiment, high
external (ecological) validity
•Avoids some participant effects
•Low in internal validity, poor control
•More time consuming
Quasi (Naturalistic) Experiment
Nature & Use
Advantages
 IV not directly manipulated
 Naturally occurring
 Participants behave naturally
 Investigates the effects of independent
variables that it would be unethical to
manipulate
Disadvantages  Participants not allocated at random to
conditions
 Difficult to identify what aspects of the
independent variable have caused the effects
on behaviour
Activity 3
Strengths & Weaknesses of Experimental Method
EXPERIMENTS
Strength
Weaknesses
Activity 3
Strengths & Weaknesses of Experimental Method
EXPERIMENTS
Strength
Weaknesses
Test hypo by manipulation of
IV
Low Ec V removed from real
life
Scientific follow standardised
procedures enable replication
Difficult to organise > sample
sizes small reduced
generalisability
Enable researcher to control
EV thereby improving
reliability of the results.
Increase anxiety and stress for
pts > ethics
Produces quantities data that
is can be SA therefore robust
and meaningful
Reductionist > simplifying
human behaviour
3. Experimental Design
Experimental: Independent Groups
Definition
Two (or more) groups of participants,
one for each condition.
•Avoids order effects
•Participants cannot guess the
purpose and of the experiment
therefore reduces demand
characteristics.
Disadvantages •Needs more participants
•Lacks control of participants
variables
Advantages
EG: LOFTUS & PALMER
Experimental: Matched Participants
Definition
Participants matched on key participant
variables
Advantages
•No order effects
•Participants variables partly controlled
•EV well controlled
Disadvantages
•Matching is difficult
EG: Bandura
Experimental: Repeated Measures
Definition
Same participants in each condition
Advantages
•Good control for participant variables
•Fewer participants
•Removes pts variables
Disadvantages
•Order effects (e.g. Boredom, practice)
•Participants guess the purpose (DC)
EG: SAMUEL & BRYANT
Activity 4
Design
1. Imagine that you are going to conduct an observation
of children in a playground with two other researchers.
To what extent to you think that you will all record the
same behaviours? How might you cope with any
disagreements?
2. If an experimenter used different wording in the
instructions to different participants, how might this
affect the results of the study?
Activity 5
Strengths & Weaknesses of Experimental Design
Strength
IM
RM
MP
Weaknesses
Activity 5
Strengths & Weaknesses of Experimental Design
Strength
Weaknesses
IM
•No order effects
•Reduce DC
•Even with random allocation
of pts to groups it is not
possible to control pts
variables as sample small
RM
•Removes Pts variables
•Order effects need to be
controlled
•Less control over task variables
as different materials need to be
presented to remove order effects
•Increase DC’s as repeat
procedure
MP
•EV are well controlled
Impossible to control of EV’s
normally matched on one/two
criterial
4. Aims & Hypothesis
Research
Definition The process of gaining knowledge
through the examination of data
derived empirically or theoretically.
Reasons for using To produce objective facts
Aims
Definition
Reasons for
using
The stated intention of a study
To be clear about the purpose of a
study
Experimental (alterative) Hypothesis
Definition
Reasons for
using
A statement of the relationship
between the IV and DV
An alternative to the null
hypothesis (accept/reject)
there will be an effect of x on y
Null Hypothesis
Definition An assumption that there is no
relationship (difference,
association, etc) in the
population from which a sample
is taken with respect to the
variables being studied.
there will be no effect of x on y
Directional Hypothesis
1 TAILED
Definition Predicts the effect/relationship
Reasons for Previous research suggests the
direction
using
Non-directional hypothesis
2 TAILED
Definition
Does not predict the direction
of the effect/relationship
Reasons for -Allows for a
using difference/relationships
occurring in either direction
-Previous research has been
inconclusive
Activity 6
Generating Hypothesis
Generate a hypothesis for each of the questions:
• What are “football hooligans” really like?
• Do children play differently at different ages?
• What are the effects of caffeine on attention and concentration?
1. Identify the independent variable (IV) and the dependent
variable (DV) from each hypothesis.
2. Identify whether your hypotheses are one tailed or two
tailed (remember one-tailed hypothesis predicts the
direction of the effect of the IV on the DV, whereas a twotailed hypothesis does not).
3. Write a null hypothesis for each of the experimental
hypotheses.
Activity 7
Null Hypothesis
Devise a suitable null and experimental hypothesis for the
following:
1. An investigator considers the effect of noise on students’
ability to concentrate and complete a word-grid. One group
only is subjected to the noise in the form of a distractor, i.e.
a television programme.
2. An investigator explores the view that there might be a link
between the amount of television children watch and their
behaviour at school.
Activity 8
Hypothesis
Read through these examples of alternative hypotheses, and
identify whether each is one- or two- tailed:
1. There will be a difference in scores on an intelligence test
between people who eat fish and those who do not eat fish.
2. There will be a relationship between extroversion and
introversion and a preference for loud music
3. People will remember more words in a foreign language if
the information is presented in picture form, rather than as
words alone.
5. Variables
Variables
Independent •Manipulated by the experimenter
Variable
•Create different conditions
Dependent
Variable
•Measures the consequence of
IV manipulation
Activity 9
IV & DV
Identify IV, DV in the following them in the following examples.
Remember:
• The IV depends on the DV
• The IV is manipulated by the experimenter or varies naturally
• The DV is one we measure
1.
2.
3.
4.
5.
Long-term separation effects emotional development more than shortterm separation. (The two variables are length of separation and
emotional development.)
Participants conform more when the model is someone they respect.
(The two variables are extent of conformity and degree of respect for the
model.)
Participants remember more words before lunch than after lunch. (The
two variables are number of words remembered and whether the test is
before or after lunch).
Boys are better than girls at throwing balls. (The two variables are gender
and ability to throw a ball).
Physical attractiveness makes a person more likeable. (The two variables
are attractiveness of a person’s photograph and whether they are rated
as more or less likeable.)
ING
NEOUS
Extraneous Variables
• Situational variables are characteristics of the environment in which
the experiment is being conducted which may have an effect on the
results. The nature of these variables is very much dependent on the
nature of the experiment but temperature, time and humidity could
all be situational variables.
• Person or Subject variables are inherent characteristics of the
Experimental Unit that might affect outcomes. Hence examples of
subject variables might include age, gender and other demographic
details (among subjects) and x, y and z (among objects) although
this is very much dependent on the object in the experiment.
• Experimental variables are characteristics of the experimenter or
the experimental team which might influence how the experiment is
conducted, or how the experimental subject responds/behaves in
the experimental setting. There is a wide definition for these
variables and they may include age, gender, qualifications, etc.
Extraneous Variables
• Situational variables : Women shown the most romantic
proposals are in a warmer room.
• Personal Variables: What if the women shown the most
romantic video clips are also more romantic in nature than the
other women?
• Experiment/Researcher Variables: What if the experimenter
was really nice to one group and he was very gruff with the
other groups?
NB extraneous variables are only important if they are present
for one group and not the other. If all of your subjects are
exposed to the same extraneous variable (like if Josh was nice
to all the subjects), then it won't change your dependent
variable and it's not considered an extraneous variable.
EXTRANEOUS Variables
Variables other than the independent variable that may bear any effect
on the behaviour of the subject being studied. Three main types
1. Subject variables: age, gender, health status, mood, background, etc.
2. Experimental variables are characteristics of the persons conducting
the experiment which might influence how a person behaves. Gender, the
presence of racial discrimination, language, or other factors may qualify
as such variables.
3. Situational variables: Air temperature, level of activity, lighting, and
the time of day.
Confounding Variables
• A confounding variable or factor is also sometimes referred to
as a confounder or a lurking variable.
• It is a "hidden" that affects the variables in question but is not
known or acknowledged, and thus (potentially) distorts the
resulting data.
• This hidden third variable causes the two measured variables
to falsely appear to be in a causal relation.
• An experiment that fails to take a confounding variable into
account is said to have poor internal validity.
Controlling EV
It is necessary to control extraneous variables so that
results are not undermined by their effect (become
confounding):
1. Control: Ensuring that an extraneous variable remains
the same for all experimental units in the experiment.
This requires that you are aware of the extraneous
variable during the design stage and that you can
control it.
2. Constant: Balance the variable across experimental
groups This enables comparisons to be made between
experimental units on the basis of the effect of the
variable.
Activity 10
Extraneous Variables
Identify the IV and DV
• Operationalise variables
• Identify EV
1. A psychologist wants to investigate whether students who
complete their 4 hours of independent study per week do
better in the psychology exam than those students who only
complete 1 hour per week...
2. An experiment to see if recall on a memory test is affected
by time of day
3. Does drinking coffee whilst revising improve exam results?
4. An experiment to investigate the effects of fatigue on
reaction time
Activity 11
Independent Variables in Core Studies
Loftus &
Palmer
Samuel &
Bryant
Bandura
Piliavin
Activity 11
Independent Variables in Core Studies
Loftus &
Palmer
•The verb in the critical question
Samuel &
Bryant
•Asking 2 questions before or after
•Materials (play-doh, water)
Bandura
•Aggressive model, non aggressive, no model
•Sex of the model
Piliavin
•Victim ‘black’ or ‘white’
•Victim ‘ill’ or ‘drunk’
•Intervention of model helper
Activity 12
Dependent Variables in Core Studies
Loftus &
Palmer
Samuel &
Bryant
Bandura
Piliavin
Activity 12
Dependent Variables in Core Studies
Loftus &
Palmer
•Estimated recalled speed
Samuel &
Bryant
•Mean number of errors in the conservation task
Bandura
•Frequency of imitative aggressive acts
•Frequency of non-imitative acts
•Frequency of verbally aggressive behaviours
Piliavin
•Time between first collapse and first helper
•No of passengers who left when victim collapsed
6. Operationalisation
‘Eating spinach affects
performance’
Operationalisation
•Variables in a form that can be tested
(operations)
•How hypothesis will be tested
Both IV and DV need to be precisely operationalised,
otherwise, the results may not be valid and cannot be
replicated.
Activity 13
Operationalisation
Here are some research ideas. For each one, identify the
IV & DV and suggest ways in which each could be
operationalised:
1. Do people remember more about a topic they are
interested in that about one in which they have little
interest?
2. Are there gender differences in the amount of aggression
shown by children in play?
3. Are neurotic people more likely to suffer from phobias?
7. Sampling Method
Population
Definition The group of people whom the sample
is drawn
Evaluation May be biased
Sample
Definition Selected to be representative of the
population
Evaluation May be biased ,therefore can’t
generalise
Random sampling
Definition
Every member of the population has
an equal chance of being selected
Advantage Potentially unbiased
Disadvantage
Most replicable
Needs to be drawn from a large
population to be unbiased
Difficult to obtain
Participants for
Psychological
Research
Volunteer Sample
Self Selected
Definition Participants become part of a study by
Advantage
volunteering
Access to a variety of participants
Ethically sound
Disadvantage Volunteer biased
Small sample
Are you available?
Opportunity Sample
Definition
Selecting people who are more easily
available
Advantage
Easy to obtain
Disadvantage
Very biased
Not replicable
Activity 14
Target Population
Identify an appropriate target population for each
project below. You would select your research sample
from this population.
1. To discover whether there are enough youth facilities
in your community.
2. To discover whether cats like dried or tinned cat food.
3. To discover whether children aged between 5 and 11
watch too much violent television.
4. To discover the causes of anxiety experienced by
participants in research studies.
Activity 15
Sampling
Find a study to illustrate volunteer sampling and another one
to illustrate opportunity sampling. (Clue: Most of the studies
you have covered used a volunteer sample, whereas some of
the studies have used opportunity samples).
1. Why do you think volunteers are more likely than nonvolunteers to be sensitive to the demand characteristics of a
study?
2. When would you not expect to find evidence of participant
reactivity?
3. Is honesty the best policy? Would demand characteristics be
reduced if both participants and experimenters knew the
true aims of the experiment?
Activity 16
Sampling in Core Studies
Stated
Y/N
Loftus & Palmer
Milgram
Maguire
Griffiths
If stated,
which one?
If not stated,
which should?
8. Factors associated with
research design
a.
b.
c.
d.
Operationalisation
Standardisation
Control of variables
Pilot studies
‘Eating spinach affects
performance’
Effects validity and replicability
Pilot study
Research is expensive (time +
money). To establish weather a
design works, that pts understand
the instructions, that nothing has
been missed out, and that pts are
able to do what is asked, a pilot study
(trial run, small scale) should be
undertaken.
Control of Extraneous Variables
Any variables that change between conditions, other than
the IV...
Experimental Control
Using techniques to ensure that extraneous variables are eliminated
Extraneous/
confounding
variables
Random Allocation
Hold constant or eliminate
Counterbalancing
Order effects balanced to make sure
each condition comes first or second in
equal amounts (ABBA)
A set of procedures that are the same
for all participants. To enable
replication.
Standardised
procedures
Participants to experimental groups;
allocate items on a test
Control of Investigator Variables (effect)
Anything that investigator does which has an effect on the participant’s
performance other than what was intended
Double blind The investigator does not know the purpose of
the experiment, to prevent expectations
influence the participant’s behaviour
Standardised A set of instructions that are the same for all
participants. To avoid investigator effects.
instructions
OR
Control of Participant Variables (effect)
Anything that has an effect on the participant’s performance other than
what was intended
Single blind Deception to prevent the participants
knowing the experimental aim
Placebo Control group thinks it is receiving the
conditions experimental treatment
Demand A demand characteristic is a subtle cue that
Characteristics makes participants aware of what the
experimenter expects to find or how
participants are expected to behave.
Standardised A set of instructions that are the same for
instructions all participants. To avoid investigator effects.
Investigator effect: Anything the investigator does which
has an effect on a participant’s performance in a study
other then what was intended.
Interviewer bias The same in an interview situation,
through, for example, leading questions
and the ‘Green-spoon’ effect
Experimenter bias The effect of an experimenter’s
expectations, communicated
unconsciously, on a participant’s
behaviour
Control Group
In the design of experiments,
treatments are applied to
experimental units in the
treatment group. In
comparative experiments,
members of the complementary
group, the control group,
receive either no treatment or a
standard treatment.
Standardisation
Procedures
Pts treated in exactly
the same way.
Instructions
Pts told what to do in
exactly the same
way.
Activity 17
Control of Variables in Core Studies
Well Controlled
Loftus &
Palmer
Samuel &
Bryant
Bandura
Piliavin
Confounding
Activity 17
Control of Variables in Core Studies
Well Controlled
Confounding
Loftus &
Palmer
Video Clips
Gender (or age)
Samuel &
Bryant
Same materials
Intelligence (or gender)
Bandura
Same toys to play with
Not matched on age
Piliavin
Same subway carriage
Participant variables
9. Reliability of Measurement
10. Validity
Generalisability
The findings of any
particular study should
apply to the whole
population
Internal
Experimental
External
Types of
Validity
Measure
Concurrent
Content
Validity: The legitimacy of a study
Internal Validity Reasons for low internal validity
The extent to which
the a measurement
technique measures
what it is supposed to
Demand Characteristics: Features of an experiment
the elicit a particular response form participants.
Participant reactivity
Extraneous variables not controlled (CV), act as an
additional IV.
Mundane realism: Do measures used generalise to
real life > contribute to external validity
External Validity Assessing external validity
Validity outside of the
research situation,
extent to which
findings can be
generalised
How representative is the sample of participants of
the population to which the results are to be
generalised? Population V
Do the research setting and situation generalise to a
real-life setting or situation? Ecological V
Do the findings generalise to the past and to the
future? Historical V
Extraneous
Internal
Mundane
Realism
Experimental
Ecological
Validity
Population
Validity
Extraneous Variables
Situational Variables
Participant Variables
Investigator Effects
Demand Characteristics
Participant Effects
External
Historical
Validity
Participant Reactivity
Participant reactivity: The fact that participants
react to cues in an experimental situation
Hawthorne Effect
Demand
Characteristics
Social Desirability
bias
Increased attention becomes a confounding
variable
Features of an experiment that a participant
unconsciously, responds to when searching for
clues about how to behave. A confounding
variable.
The desire to appear favourably
Validity of Psychological Measure
Concurrent How well does the measure agree with
Validity existing measures?
- Test using old and new tests
Content Validity Does the method used actually seem to
measure what you intended?
- Use a panel of experts
Concurrent
Measure
Content
Activity 18
Ecological Validity in Core Studies
How/Low
Loftus & Palmer
Samuel & Bryant
Dement & Kleitman
Milgram
Evidence
11. Ethical Issues
11. Ethical Issues
 Deception
 Informed consent
 Psychological harm
 Informed consent
 Difficulty debriefing
 Privacy
 Confidentiality
11. Dealing with Ethical Issues
Presumptive…
A
Prior…
B
C
Dealing with informed consent
• Presumptive consent: Ask for others’ opinion
and presume participants feel the same way.
• Prior general consent: Get participants to
agree to take part in a number of studies, one
of which they will be deceived in.
11. Dealing with Ethical Issues
Dealing with deception
• Debriefing: Inform participants of true nature
of the study after it is conducted and allow
them to discuss their feelings.
• Right to withhold information
• Cost and benefits: Deception is acceptable if
the benefits are sufficient.
11. Dealing with Ethical Issues
Dealing with protection from
psychological harm
• Anticipating harm and stopping
• Using role-play
• Use of questionnaires: Ask people how they
would behave.
• Debriefing
Activity 19
Ethical Issues in Core Studies
Informed Deception
Consent
Loftus &
Palmer
Samuel &
Bryant
Dement &
Kleitman
Milgram
RTW
Debrief
12. Data
a. Type of data
b. Descriptive Statistics
c. Inferential Statistics
Types of Data
Data Collection
Quantitative
Data
Qualitative
Data
Easy to analysis
 Oversimplifies
reality
Produces neat
conclusions
Represents the
complexity of human
behaviour
More difficult to
detect patterns and
reach conclusions
Provides rich data
Subjective, affected
by personal
expectations and
beliefs
Descriptive Statistics
1. Measures of central tendency
2. Measures of dispersion
3. Graphical representation
2,4,4,5,6,6,7,7
8,8,8,8,8
9,10,11,11,12
Measures of Central Tendency
Mean: Add  Makes use of all the Can be
values, divide by data
misrepresentative
number of
if there are
values
extreme values.
Median: Middle  Not affected by
value in an extreme scores
ordered list
Not as ‘sensitive’
as the mean
Mode: The most  The mist common
common value(s)
value(s)
Not useful when
there are several
modes
Measures of Dispersion
• Measures of
dispersion
• the range
• standard deviation
Measures of Dispersion
Range Highest to
Easy to calculate
lowest
Affected by extreme
values
Standard SD measures
Precise, all values
Deviation the amount of taken into account
variation or
dispersion from Harder to calculate
the average.
Graphs & Charts
Histogram Graph showing continuous frequency data
with a true zero e.g Exam results 0-30marks
Bar Charts Graph showing frequency data; data need not
be continuous e.g. Categories
Scattergraph For correlations. Scatter of dots; each dot
represent one case
Inferential Statistics
1.
2.
3.
4.
Level of data
Levels of significance
Tests
Type 1 & Type 2 Error
Inferential Statistics
June 2012:
• What does p ≤ 0.05 level of significance mean? (2)
• This means that this is a 95% probability that the change in the DV is
as a result of the IV and a less than 5% probability that this is due to
random chance. The findings are therefore statistically significant.
Thus we can reject the null hypothesis and accept the
experimental/alternative hypothesis.
• If you obtained this level of significance in your practical project,
explain what this would mean in relation to your null hypothesis. (4)
June 2011:
You must use a repeated measures design experiment and plan to
collect quantitative data. State an appropriate statistical test to
analyse the data you would collect. Give reasons for your choice. (3)
Inferential Statistics
June 2010:
You must use a correlation design and plan to collect at least
ordinal data. State an appropriate statistical test (nonparametric) to analyse the data you would collect. Give reasons
for your choice. (3)
June 2013: How could you obtain nominal level data from your
practical project? (3)
June 2014: You must use an independent measure design and
plan to collect data which measures observable behaviour. State
and appropriate inferential statistical test to analyse the data
that would be collected in your practical project. Give reasons
for your choice. (3
Inferential Statistics
Specimen Papers:
• You must use a repeated measures design experiment and
plan to collect at least ordinal level data. State an appropriate
statistical test to analyse the data you would collect. Give
reasons for your choice. (3)
• If having carried out your investigation and an inferential
statistical test, your experimental hypothesis was found to be
significant for p≤0.05, what would p≤0.05 mean? (3)
• Identify and describe the rational for choosing an appropriate
inferential statistical test that could be used to calculate the
significance of any correlation in your practical project. (3)
• If your practical project comes up with the findings or what
does p ≤ 0.05 level of significance mean? (2)
Inferential Statistics
Experiment
• This means that this is a 95% probability that the change in the DV is as a result
of the IV and a less than 5% probability that this is due to random chance. The
findings are therefore statistically significant. Thus we can reject the null
hypothesis and accept the experimental hypothesis.
Non-Experimental
• This means that this is a 95% probability that the change in the DV is as a result
of the IV and a less than 5% probability that this is due to random chance. The
findings are therefore statistically significant. Thus we can reject the null
hypothesis and accept the alternative hypothesis.
Correlation
• This means that this is a 95% probability that the relationship between the two
variables is significant and a less than 5% probability that this is due to random
chance. The findings are therefore statistically significant. Thus we can reject
the null hypothesis and accept the alternative hypothesis.
Nominal Data
Ordinal Data
Interval/Ratio Data
Levels of Significance
Is there a difference between the groups that is REAL or is it
due to CHANCE?
Level
Probability Significance
When used
1%
(p<0.01)
Highly
Where we would want to take few chances
5%
(p<0.05)
Significant
Acceptable for psychological research
10%
(p<0.10)
Marginally
May indicate need for better methodology
Less than
Inferential statistics
Non-Parametric statistical tests:
1. Mann Whitney U: Test significance of the difference between
two conditions when an independent design has been used
and the level of data is at least ordinal.
2. Wilcoxon matched participant signed rank: Test significance of
the difference between two conditions when an repeated
measures design has been used and the level of data is at least
ordinal.
3. Chi-Squared: test of significance of association used when
nominal data has been collected.
4. Spearmans Rho Test: when data is at least ordinal and a
correlational method is used.
Level of
measurement
Categorical
Nominal
Difference/Association
Repeated
Independent
Sign Test
Chi Squared
Correlation
Measured
Ordinal
Wilcoxon T
Mann Whitney
(matched pairs)
U
Spearman
(rho)
Type 1 & 2 Errors
There is a possibility that errors may have been made:
Type 1 Error Deciding to reject the null when actually the
results was due to chance or some other factor.
Type 2 Error Deciding to retain the null when actually the
result was caused by the IV.
Too Low
P = <0.10
Type 1 Error more likely
Too High
P = <0.001
Type 2 Error more likely
13. Improvements & Future Research
Improvements
Indentify a limitation and
suggest how you make
improvements to the
investigation:
Future Research
1.
•
•
Sample
Method
Validity
Reliability
Ethics
Think about changes that you could
make to your stimulus
Could you use a different method?
2.
What implications does your
research have on society? What
group would benefit from
knowing this information
(application)
3.
Does this study have real life
validity?
Example:
•
•
•
•
•
If you were to progress research
in this area what would you do
next?
14. Writing
procedures…
14. Writing a procedure for Experiments
1. Decide aim, research question and alternative/null
hypothesis
2. Plan procedure, including obtaining ethics approval,
choosing experimental design, operationalising the
hypothesis, preparation of materials and deciding sampling
method.
3. Obtain sample and make arrangements for conduct of study
4. Obtain informed consent from participants
5. Allocate participants to experimental conditions and give
instructions.
6. Participants follow experimental steps (data collected)
7. Thank and debrief participants
8. Analysis data, produce conclusions
9. Write report of practical investigation
14. Writing a procedure for Observations
1. Decide aim and research question
2. Plan procedure: obtain ethics approval, draw up schedule (if
structured observation): choose and train observers: plan
time and location for observation.
3. Possibly run pilot study in order to check on usefulness of
selected categories and feasibility.
4. Covert observations – no informed consent/overt Obs – pats
informed that observation will be conducted.
5. Place observers in position.
6. Conduct observation – pts are observed for designated
period while observers record behaviours (data collected).
7. Thank and debriefed pts (overt)
8. Analyse data, produce findings and draw conclusions
9. Write report of practical investigation.
Qualitative Methods
Observations
Self Report
Correlation
Correlational Analysis
Nature & Use
Advantages
Co-variables examined for positive, negative or zero
association.
•Allows study of hypotheses that cannot be
examined directly
•More data on more variables can be collected
more quickly that in an experiment set-up
•Problems of interpretation are reduced when no
association is found
Disadvantages
•Interpretation of results is difficult
•Cause and effect cannot be established
•Direction of causality is uncertain
•Variables other than one of interest may be
operating
Ethical Issues
Misunderstanding of findings because causal
relationship cannot be established
Questionnaire
Nature & Use
Set of questions
Design: open (qual) or closed (quan)
Advantages
•A lot of data can be collected
•Does not require specialist administrators
•People are more willing to answer embarrassing
or personal questions on an anonymous written
questionnaire than in a face-to-face interview
Disadvantages
•Leading questions, social desirability bias
•Biased samples
Ethical Issues
Confidentiality
Privacy
Interview
Nature & Use
Questions can be predetermined, or created in
response to answers.
Design: structured/ unstructured questions
Advantages
•Lots of ‘rich’ data
•Telephone interviews
Disadvantages
•Social desirability bias, interview bias
•Requires skilled personnel
Ethical Issues
Confidentiality
Privacy