CW2 – Test
Section 1: Reliability & Validity
(1)Reliability:
a. Extent to which something is consistent
b. Reliable measures are those that produce consistent
results over time.
c. Reliable effects are those that can be replicated.
(2)Validity:
a. Extent to which something is well-founded
b. Valid measures are those that measure what they claim to
measure
c. Valid effects are those than can be attributed to IV
Systematic bias – validity
- Participant bias
o Hawthorne effect – pp behaviour altered through taking
part in experiments = pp nervous or demand
characteristics
o Demand characteristics – try to work out the aim of study
o Evaluation apprehension
o Selection bias (sampling)
- Experimenter bias
o E.g attitue and behaviour towards pp
o Experimenter expectancy
o Uniinentional non-verbal cues
o Flawed experimental design or statistical analyses
Section 2: Error and Power
, (1)Type-1 error
a. Accepting the alternative hypothesis when the null is
correct
b. Conclude that the result is significant when it isn’t
(2)Type-2 error
a. Accept null hypotheses when the alternative is true
b. Conclude that a result is non-sig when it is
c. Power is the likelihood of not making a type-2 error
Effect size & likelihood ratio
(1)Effect sizes
a. Statistics of the size of our effect
b. Large enough to be of practical use?
(2)Types of effects:
a. Eta-sq
b. Cohen’s d
Likelihood ratios – tell us how likely our alt. hypothesis is in
comparison to null hypothesis.
Section 3: Questionnaires
(1)Good questionnaires
a. Differentiate ppl who differ on the trait of interest
b. Standardized
c. High levels of internal and external reliability
d. Be valid measures of the variable of interest
(2)Things to consider:
a. Item construction
b. Item organistion (sequencing; reverse scoring)
c. Response format
2
Section 1: Reliability & Validity
(1)Reliability:
a. Extent to which something is consistent
b. Reliable measures are those that produce consistent
results over time.
c. Reliable effects are those that can be replicated.
(2)Validity:
a. Extent to which something is well-founded
b. Valid measures are those that measure what they claim to
measure
c. Valid effects are those than can be attributed to IV
Systematic bias – validity
- Participant bias
o Hawthorne effect – pp behaviour altered through taking
part in experiments = pp nervous or demand
characteristics
o Demand characteristics – try to work out the aim of study
o Evaluation apprehension
o Selection bias (sampling)
- Experimenter bias
o E.g attitue and behaviour towards pp
o Experimenter expectancy
o Uniinentional non-verbal cues
o Flawed experimental design or statistical analyses
Section 2: Error and Power
, (1)Type-1 error
a. Accepting the alternative hypothesis when the null is
correct
b. Conclude that the result is significant when it isn’t
(2)Type-2 error
a. Accept null hypotheses when the alternative is true
b. Conclude that a result is non-sig when it is
c. Power is the likelihood of not making a type-2 error
Effect size & likelihood ratio
(1)Effect sizes
a. Statistics of the size of our effect
b. Large enough to be of practical use?
(2)Types of effects:
a. Eta-sq
b. Cohen’s d
Likelihood ratios – tell us how likely our alt. hypothesis is in
comparison to null hypothesis.
Section 3: Questionnaires
(1)Good questionnaires
a. Differentiate ppl who differ on the trait of interest
b. Standardized
c. High levels of internal and external reliability
d. Be valid measures of the variable of interest
(2)Things to consider:
a. Item construction
b. Item organistion (sequencing; reverse scoring)
c. Response format
2
