Basics of Experimental Design Flashcards
1
Q
What is a fact?
A
A statement about a direct observation of nature that is so consistently repeated that virtually no doubt exists to its truth value
2
Q
What is a theory?
A
A collection of statements (propositions, hypothesis) that together attempt to explain a set of observed phenomena (e.g. evolution)
3
Q
What is a hypothesis?
A
A clear but tentative explanation for an observed phenomenon (something that needs to be tested and proven enough)
4
Q
What are features of theories?
A
They are integrated set of proposals that:
- define
- explain
- organise
- interrelate
Theories are proposals that provide a model of how the observed phenomena work and make general predictions upon which specific hypotheses can be based.
5
Q
Hypotheses make specific predictions. What must they be?
A
- Falsifiable
- Testable
- Precisely stated: are all terms clearly defined?
- Rational: is it consistent with known information?
- Parsimonious: Is the explanation the simplest possible.
6
Q
What are constructs?
A
- Building blocks of theories
- Theoretical concepts formulated to serve as causal or descriptive explanations
- don’t directly indicate a means by which they can be measured
7
Q
What are variables?
A
- any characteristic that can assume multiple values (e.g. gender, body weight)
- An event or condition the researcher observed or measures
- Variables must be operational (i.e. explicitly stated)
8
Q
What are the different scales of measurement?
A
- Nominal (category membership)
- ordinal (ranked or ordered)
- Interval (equal increments but no real 0 point)
- Ratio (real 0 point)
- These are organised in a hierarchy with ratio giving us the highest level of data
9
Q
What is nominal (categorical) data?
A
- category membership
- numbers assigned serve as labels but do not indicate numerical relationship
- E.g. gender, political party, religion
10
Q
What is ordinal data?
A
- data can be ranked along a continuum
- intervals between ranks are not necessarily equal
- e.g. running race positions, attractiveness
11
Q
What is interval data?
A
- intervals between successive values are equal
- no true ‘zero’ point (no absence of something)
- e.g. temperature, shoe size
12
Q
What is ratio data?
A
- highest level of data
- equal intervals and a true zero point
- e.g. height, distance, time
13
Q
What are independent variables?
A
- The variable that is manipulated and is hypothesised to bring about a change in the variable of interest
- also known as the grouping variable
- has at least 2 levels (conditions)
14
Q
What’s the dependent variable?
A
- The variable that is measured (a.k.a. the outcome variable)
- We compare differences in the DV under the different levels of the IV.
- E.g. exam score
- E.g. score on a test of intelligence
- E.g. score on a test of mood
- Independent variable -> affects change -> dependent variable
15
Q
What is subjects design?
A
The assignment of participants to experimental conditions (levels of the IV)
16
Q
What are the different types of subject design?
A
- Between subjects/independent groups
- Within subjects/repeated measures
- Mixed-designs (mixture of between and within)
17
Q
What are between subject designs? (independent groups)
A
- Participants each exposed to one level of the IV
- Experimenter assigns participants to one of the groups
- E.g. Alcohol consumption of short term memory
- IV: Alcohol consumption
- DV: Memory performance
- Assign participants to one of two groups (alcohol or no alcohol)
- Administer alcohol accordingly
- Measure each group’s memory performance and compare
18
Q
What are within subject designs? (repeated measures)
A
- Participants exposed to all levels of IV
- E.g. Alcohol consumption on short term memory
- IV: alcohol consumption
- DV: memory groups
- Participants now take part in both levels of IV (test before alcohol and test after alcohol
- Measure each Ps performance before and after alcohol and compare
19
Q
What are the considerations with Between Subjects designs?
A
- Participants are all inherently different which may affect outcome (e.g. some may be tired on that day or some may have an ADHD diagnoses)
- We can’t eliminate the effects of these other variables
- But we can minimise these effects by spreading their influence across the different levels of our IV
20
Q
When is random allocation used and what does it do?
A
It is used in between subjects designs and it ensures each participant is likely to be assigned to any IV level
21
Q
Why should you use random allocation?
A
Distributes the occurrence of potential moderating variables equally among experimental conditions
Prevents experimenters (un)intentionally biasing their results
Enables the use of powerful statistical tests that can help determine causal relationships between variables
22
Q
What are the considerations with within-subjects designs?
A
- Potentially moderating characteristics are kept equal across the levels of the IV (each participant acts as his/her own control)
- Requires fewer participants
- Problem of Order effects: Once participants have been exposed to one level of the IV there’s no way to return them to their original states
23
Q
What should you use to minimise the effect of order effects in within-subjects designs?
A
Counterbalancing
24
Q
How does counterbalancing work?
A
- Split the group of participants in half (A and B)
- Group A can participate in level 1 then 2
- Group B can participate in level 2 then 1
- Order effects will still influence Ps performance but the effect of that influence will be evenly spread across each level of the IV
25
What are factorial designs
Experimental designs with 2 or more IVs
26
What do factorial designs allow us to ask?
- What effect does IV1 have on the DV
- What effect does IV2 have on the DV
- What effect does the interaction of IV1 and IV2 have on the DV
27
Give an example of a factorial design
- Effects of alcohol consumption and work shift patterns on work productivity:
- DV: work output
- IV: shift pattern
- IV: alcohol consumption
28
How does a full independent factorial design (between subjects) work?
- E.g. Effects of alcohol consumption and work shift patterns on work productivity:
- DV: work output
- IV: time of day – between subjects
- Randomly assign Ps to either the day shift of night shift
- IV: alcohol consumption – between subjects
- Randomly assign Ps to either take the test with or without alcohol
- Each participant takes part in just one experimental condition
29
How does a fully repeated measures factorial design (within subjects) work?
- E.g. Effects of alcohol consumption and work shift patterns on work productivity:
- DV: work output
- IV: time of day – within subjects
- Each groups takes the test during a day shift and a night shift
- Counterbalance to control for order effects
- IV: alcohol consumption – within subjects
- Each group takes the test with alcohol and without alcohol
- Counterbalance to control for order effects
- Each participant takes part in all experimental conditions
30
How do factorial mixed designs work?
- E.g. Effects of alcohol consumption and work shift patterns on work productivity:
- DV: work output
- IV: time of day – between subjects
- Randomly assign Ps to either the dayshift or nightshift groups
- IV: alcohol consumption - within subjects
- Each group takes the test with and without alcohol
- Counterbalance
- Always contains
1 or more within subjects IV(s)
1 or more between subjects IV(s)
- Each participant takes part in all levels of within subject IV(s) but just one level of between subject IV(s)
31
What does choice in experimental design depend on?
whether you are more concerned about getting rid of order effects (between subjects) or getting rid of individual difference effects (within subjects) and the number of participants you have available
32
What are quasi-experimental designs?
the assignment of participants to experimental conditions is pre-determined. This is because occasionally it's not possible to randomly assign Ps to the levels of the IV as they may be a pre-existing group (e.g. alcoholics) or based on fixed characteristics (e.g. gender)
33
Why do quasi-experimental designs pose a problem?
there are likely to be differences between the groups other than the variable of interest as you can not use random allocation
34
What do you do to overcome the problem of quasi-experimental designs?
Matching groups:
- identify potentially moderating variables and match the groups on this basis
Matching pairs:
- matching pairs on the basis of variables such as educational background ect. is idea but it is usually impossible to find perfectly matching Ps
35
Occasionally you can’t counterbalance the order in which Ps are exposed to the levels of the IV. For example examining the effectiveness of mnemonic training on memory performance. Why does this pose a problems?
- There are likely to be differences between time 1 and time 2 than the variable of interest
- once the P knows what the experimenter 'wants' them to do they might not act as they naturally would
36
How do you overcome the problem of not being able to counterbalance with within subject designs?
pre-test post-test control group design
37
What is a pre-test post-test control group design
Split Ps into 2 groups and manipulate the IV of interest in one group only
Treatment group -> test -> manipulation -> test
Control group -> test – no manipulation -> test
- The inclusion of a control group allows us to account for any order effects that may be present
- We can then statistically control for the difference in the treatment group accounted for by order effects
38
What are the terms used to describe between subjects and within subjects designs in developmental research?
- Between subjects = cross-sectional
| - Within subjects = longitudinal
39
What is dichotomising?
dividing into two categories of continuous and discrete variables
40
What are extraneous variables?
factors that the researcher might not have accounted for but might influence the outcome. These variables need to be controlled
41
What is a confounding variable?
specific type of extraneous variable that is related to both of the main variables that we are interested in.
42
What are demand effects?
when in within-participants design the participants realise what the experimenter wants them to do and perform on how they believe they should do rather than how they would actually act
