PSYC125 Red Topics

?
  • Created by: Caitlinl
  • Created on: 27-05-17 18:49
What is statistical power?
The probability you find a statistically significant difference, provided it exists
1 of 105
What is incomplete within subjects design?
All possible combinations; done via factorial; only usually done with 3 or less groups/conditions
2 of 105
What is selected orders (Latin Square Design)?
Equally often in each position; has to be a multiple of the number of conditions; each condition precedes and follows only once
3 of 105
What is a disadvantage of the Latin Square Design?
Individual differences can still be seen
4 of 105
What is the complete within subjects design?
Each condition administered several times in different orders
5 of 105
What is an advantage of complete within subjects design?
Balance practice effects as each participant completes all conditions
6 of 105
What are the counterbalancing methods?
Block randomisation; Does all possible orders
7 of 105
What is the advantage of complete within subjects design?
Practice effects diminished
8 of 105
What is the ABBA Design?
Presents 1 random sequence of conditions; then present opposite; same amount of practice effects
9 of 105
What are the disadvantages of the ABBA Design?
Only balances linear practice effects; no use when non-linear
10 of 105
What are Linear Practice Effects?
Participants gain 1 "unit" of practice with each presentation of the condition
11 of 105
What are Non-Linear Practice Effects?
Participants work out order and anticipate next
12 of 105
What are the disadvantages of Non-Linear Practice Effects?
No use testing of ind diffs; time consuming; levels of IV are present an unfolding time sequence that can't be undone; can't do it when can't repeat DV; Differential Transfer Effects
13 of 105
What are differential transfer effects?
The effects of the condition before affecting the condition after
14 of 105
What is sensitivity?
Detecting small effect of IV on DV
15 of 105
What is the floor effect?
Low scores because the task is too hard (no room for variation)
16 of 105
What is the ceiling effect?
High scores because the task is too easy (no room for variation)
17 of 105
What is the solution to overcoming the floor/ceiling effect?
Not make the tasks too hard or too easy
18 of 105
What do you have to consider in regards to the sample?
Appropriate to the task
19 of 105
What are the 3 criteria for internal validity?
Show co-variation/correlation; show time/order relationship; eliminate all other possible causes
20 of 105
What are the 5 key threats to internal validity?
Test-intact groups; order/practice/fatigue; differential transfer effects; extraneous variables; unequal loss across groups
21 of 105
What are in tact groups?
Between subjects design
22 of 105
How to practice/order/fatigue effects occur?
Within subjects designs; (practice get better, fatigue get worse)
23 of 105
How do we balance these effects?
Counterbalancing
24 of 105
What are extraneous variables?
Cause variation not related to manipulation; e.g. room, temperature, experimenter etc
25 of 105
What is unequal loss across groups?
Ppts dropping out of one condition significantly more than the other (only in between subjects design)
26 of 105
What are participant expectancy effects?
E.g. cannabis and memory - expect to do worse, so do worse
27 of 105
How do you get over participant expectancy effects?
Use a placebo
28 of 105
What are experimenter expectancy effects?
Experimenter affects the outcome of the experiment
29 of 105
What is a study highlighting experimenter expectancy effects?
Rosenthal and Fode (1963)
30 of 105
What did Rosenthal and Fode do?
Altered the lighting in a work place to see if it affected work efficiency, workers copped on and whenever it changed worked harder, as knew they were being observed
31 of 105
How do you get around this?
Double-blind trials
32 of 105
What is external validity
Applying the findings to real life
33 of 105
What is convergent validity?
The multi-modal approach; conduct controlled experiments and a naturalistic study, if you come to the same conclusions you have convergent val
34 of 105
What are the assumptions for t-tests?
Parametricity of the DV and it meets parametric assumptions
35 of 105
How are these aims achieved?
Interval/ratio data (only ordinal if have a massive sample); report mean and SDs, have 95% confidence intervals
36 of 105
What do you then do to work it out?
Multiply the standard error by +/- 1.96 to give you 95% confidence levels
37 of 105
What happens if the sample mean is similar (small difference)?
Same population= selected similar groups, different population = similar means
38 of 105
When can t tests be used?
Sampling different populations with similar means
39 of 105
What happens when sample means are dissimilar (or a large difference?)
Same population and randomly selected extremes; different populations with different means
40 of 105
What is the sample Standard Deviation?
Standard Error comparing sample means (mean difference divided by standard error)
41 of 105
What are the main features of t tests?
Powerful test of difference; divide mean of difference by standard deviation of difference of scores; same as linear regression models
42 of 105
Should we have variance?
Yes and we can have it for variety of reasons
43 of 105
What are these 2 reasons?
Our model or anything else
44 of 105
What is total variance?
Model variance + error variance
45 of 105
What is manipulation variance?
How far the scores are from the grand mean
46 of 105
What is error variance?
How individual scores differ from the group mean
47 of 105
What do you need to know for the anatomy of equations?
Total sum of squares; manipulation sum of squares; error sum of squares and degrees of freedom
48 of 105
How do you report an unrelated ANOVA?
An unrelated one way ANOVA found a (significant) effect of X on Y, F(critical value, df) = _, P(
49 of 105
What do we not know when conducted an ANOVA?
If there is an overall effect
50 of 105
How do we address this?
By conducting post-hoc analysis (planned comparisons)
51 of 105
What do we want to try and find out?
If comparing a placebo has significant differences; (sig dif between control and any intake); compare high dose to other level
52 of 105
What are the features of post-hoc analysis?
No a priori assumptions; Compare all groups to each other; decide which tests to use after initial analysis
53 of 105
What are the features of planned comparisons?
A priori assumptions; few meaningful comparisons; decide on tests before initial analysis
54 of 105
What are the features of Scheffe's test?
All possible comparisons
55 of 105
How do you conduct Scheffe's test?
Compare averages of groups A & B to averages of groups C, D and E
56 of 105
What are the features of pairwise comparisons?
All possible pairwise comparisons
57 of 105
How do you conduct Pairwise comparisons?
Compare the mean of every group with the mean of every group with the mean of every other group
58 of 105
What is Dunnett's test?
All groups compared with a control group and compare all groups to this, not each other
59 of 105
What is Bonferroni's test (NOT CORRECTION)
Compare specific groups based on theoretical comparisons e.g. only comparing A and B or B and C
60 of 105
What should planned comparisons always be?
Theoretically sound!!
61 of 105
What do you have to remember about planned comparisons?
The more comparisons, the higher the chance of making a type 1 error
62 of 105
How do you apply Bonferroni's correction?
Divide the accepted level of significance (P
63 of 105
Do less planned comparisons have more or less statistical power than loads of unplanned post hoc tests
More as controlling for type 1 errors
64 of 105
What can Student Newman Keuls test?
All comparisons
65 of 105
How does SNK work?
Works on difference sums (means); find the largest difference between mean groups and work down
66 of 105
What happens next?
Have all possible differences between all group means (observed values); difference should be greater than critical value
67 of 105
What parameters should you use?
Degrees of freedom as given in main ANOVA
68 of 105
What are the assumptions of an ANOVA?
Interval/ratio level data; normally distributed; homogeneity of variance; homoscedasticity and sphericity
69 of 105
What are the stoic principles underlying qualitative research?
Aristotle's tabula rasa - born a blank slate; knowledge enters through the senses; truth obtained through reason
70 of 105
What are the sophist principles underlying qualitative research?
No such thing as absolute truth; knowledge is individual opinion; truth obtained via discourse
71 of 105
What are the empiricist principles underlying qualitative research?
John Locke - knowledge is experience vis the senses; experimentation & observation; scientific approach to study of the mind
72 of 105
What are the Anthony Ashley-Cooper's principles underlying qualitative research?
Embrace the social; look at behaviour in a socio-historical context; scientific truth comes from discourse
73 of 105
What are the Wundt's principles underlying qualitative research
Founder of modern psychology; started qualitative research; incorporated culture, expression, beliefs, morality and imagination
74 of 105
What are the William James' principles underlying qualitative research?
Radical empiricism; objective analysis; mind, experiences and nature are inseperable
75 of 105
What are the scientific empiricist views?
Quantifying individual experiences; atomist point of view (single our brain functions)
76 of 105
How can change occur?
War, unethical experiments, feminism and increased world funding
77 of 105
What is the post-modernist view of qualitative research?
No single clear window into the mind; filters affect behaviour; wide range of methologies; ongoing debate
78 of 105
What are the features of qualitative research?
Qual data, natural, focus on meaning, cultural patterns; description; idealistic
79 of 105
What are the contrasting features of quantitative research?
Quant data; artificial; focuses on behaviour; seeks scientific laws; prediction; realistic
80 of 105
What are the advantages of quantitative research?
Purity of data and control; More order (theory building); reliable, valid and generalisable; strong roots in behaviourism
81 of 105
What are the disadvantages of quantitative research?
Stripped of context; meaning and purpose detached; Responses don't help us understand the individual; specific contexts; limited analysis
82 of 105
What are the advantages of qualitative research?
Rich in detail; examine underlying structures of data; not as strict interpretations; context and meaning included
83 of 105
Why is it useful to combine qual and quant research?
Useful 1st step in research; both unobtrusive and obtrusive
84 of 105
What is unobtrusive data collection?
No contact/influence on data collected; non-reactive; data usually exists before research; greater value in some fields (criminality e.g.)
85 of 105
What is acretion?
Build up of physical traces
86 of 105
What are some examples of acretion?
Litter, graffiti; latrinalism; art
87 of 105
What is does erosion indicate?
Desire lines; flaws in public space designs; social relationships; wear on flooring; wear and tear of a book (indicating popularity)
88 of 105
What are the advantages of erosion?
Tells us about actions; high face validity; data is readily available
89 of 105
What are the disadvantages of erosion?
Intention inferred; additional evidence needed; bias- degrade over time
90 of 105
What does archival data consist of?
Running records (databases); doctors records; police records
91 of 105
What are examples of episodic/discontinous/personal records?
Police interviews; witness statements; company/institution records; media; diaries
92 of 105
What is the disadvantage of episodic records
Bias - Selective deposit; selective addition/omission; selective survival
93 of 105
What is simple observation?
Unobserved, passive, unobtrusive role; no control over behaviour
94 of 105
What techniques are used in simple observation?
Notes, audio recording; video recording; checklists
95 of 105
What are the 5 types of simple observation according to Lee, 2000?
Exterior physical signs; expressive movements; physical movement; in-situ conversations and time related behaviour
96 of 105
What are the advantages of simple observation?
Can't be realistically/ethically reproduced in a lab; chronology of events; accurate permanent records (can be reanalysed in the future)
97 of 105
What are the disadvantages of simple observation?
External validity - subjective analysis; selective recordings; assumptions about human behaviour made - social rules
98 of 105
What are the challenges of unobtrusive data collection?
Not true research data; recorded for other reasons; low quality and validity; ethical breaches; hard to get access; bias; doesn't tell us why
99 of 105
What strategies can be employed to deal with this?
Consider data origins; standardise recording procedures; be alert to bias; corroboration of different sources; take care in drawing inferences
100 of 105
What are the features of obtrusive data collection?
Generated for current research; take the form of field studies, focus groups; researcher involved with participants
101 of 105
What is discourse analysis?
How language constructs social reality and social contact
102 of 105
What is included?
All forms of texts and speech
103 of 105
What is performative discourse?
Produces phenomena, it regulates and constrains
104 of 105
What is functional discourse?
Provides a means to an end - sophist discourse - social structure
105 of 105

Other cards in this set

Card 2

Front

What is incomplete within subjects design?

Back

All possible combinations; done via factorial; only usually done with 3 or less groups/conditions

Card 3

Front

What is selected orders (Latin Square Design)?

Back

Preview of the front of card 3

Card 4

Front

What is a disadvantage of the Latin Square Design?

Back

Preview of the front of card 4

Card 5

Front

What is the complete within subjects design?

Back

Preview of the front of card 5
View more cards

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Statistics & Research Methods resources »