Experimental and Quasi-experimental methods

Question 1

How does true experimental research methods differ from correlational studies?

Answer 1

Experimental methods test for the presence of cause and effect.
Other methods generally seek to reveal relationships between variables.
After an experiment we can conclude more than a relationship, one variable may directly affect another

Question 2

What is statistical power?

Answer 2

Statistical power is a measure of the likelihood that a researcher will find statistical significance in a sample if the effect exists in the full population. Power is a function of three primary factors and one secondary factor: sample size, effect size, significance level, and the power of the statistic used.

Question 3

How can we conclude cause and effect?

Answer 3

By controlling potential sources of variance

Question 4

What is the simplest experimental design?

Answer 4

random selection and allocation
experimental group - receives treatment
control group - does not receive treatment

Question 5

Name the two types of pre-experimental designs and how do they differ to experimental designs?

Answer 5

1. One-shot case study
2. One Group pretest post test

they differ in that they do not have random assignment and no control group

Question 6

Name the three basic Experimental research designs?

What are their basic characteristics?

Answer 6

1. Pretest Post Test Control Group
2. Post Test Only Control Group
3. Solomon 4 Group

Their basic characteristics are random assignment and a control group

Question 7

What are some disadvantages for the Post Test Only Control Group Design?

Answer 7

Pretest doesn’t occur so assist in assigning to groups

If randomisation is not effective, groups may not be equal

Question 8

What is an RCT?

Answer 8

Randomised Control Trial
Used in clinical settings to test a new treatment. It is a type of intervention study - participants receive an intervention or treatment (or not) to evaluate its’ effectiveness on typically health related outcomes.
Considered gold standard for clinical trials

Question 9

What are some key characteristics to intervention studies?

Answer 9

participants are randomly allocated to treatment or control group
randomisation minimises selection bias
RCT considered gold standard for clinical trials

Question 10

What is regression?

Answer 10

Regression - using linear relationship to predict the average numerical value of one variable given the value of another (using a straight line - ie regression line). eg you can predict the average of Y from X.
If you establish at least a MODERATE correlation between X and Y through correlation coefficient and scatterplot, you know they have some form of linear relationship.

Question 11

What are extraneous variables?

Answer 11

Extraneous variables are unwanted variables that influence the relationship between the variables being examined. They need to be controlled in order to prevent them from confounding the results.

Question 12

What is confounding?

Answer 12

Confounding occurs when the extraneous variable varies systematically across the levels of the independent variable, and is correlated with the dependent variable

Question 13

What is an example of confounding variables?

Answer 13

Different temperature in different testing conditions

Different mean age in different groups

Question 14

How can we control for extraneous variables?

Answer 14

1. Matching - ensures subjects in each group are equivalent on some characteristic (albeit time consuming and costly)
2. Homogenous Groups - select sample where critical factors are alike (ethnicity, education, political attitude)
3. Randomisation - random selection and assignment and random order of presentation of experimental conditions
4. Analysis of Co-variance

Question 15

What is Counterbalancing?

Answer 15

Counter-balancing - controls for order effects. A procedure for controlling order effects by presenting conditions in different sequences

Question 16

What is Analysis of Covariance?

Answer 16

Statistical tool that equalises or adjusts any initial differences that may exist between groups on a covariate, and subtracts the influence of the relationship between the covariate and the dependent variable from the effect of one treatment

Question 17

What is a covariate?

Answer 17

Nuisance or concomitant variables. Sometimes continuous variables.
A variable that is observed in a statistical experiment, but is not specifically measured or utilized in the analysis of the data. It is sometimes necessary to correct for concomitant variables in order to prevent distortion of the results of experiments or research.

Question 18

What are the advantages of a true experiment?

Answer 18

Accurate measurement of variables
control
compare and analyse data
easier to replicate

Question 19

What are the disadvantages to a true experiment?

Answer 19

variables may be narrowly defined (construct validity)
lack of generalisation beyond experimental setting
artificial nature - may lead to artificial results

Question 20

What is the key difference between Quasi Experiments and True Experiments?

Answer 20

Random assignment of participants is not possible in quasi-experimental methods, and useful when full experimental procedures are not possible for ethical practical or logical reasons.

Question 21

If random assignment cannot be undertaken and there is NO control group - what type of design is this?

Answer 21

Pre-experimental

Question 22

If random assignment cannot be undertaken and there IS a control group - what type of design is this?

Answer 22

Quasi-experimental

Question 23

What are some types of Quasi-experimental designs?

Answer 23

1. Non-equivalent control group design (when random assignment is not possible, similar to pretest/posttest control group but no random selection or assignment - ie selection bias…)
2. Static group comparison (no pretest and only used when there’s no other option)
3. Single Subject Design (used in behavioural analysis and special education. Has options - AB, ABA and ABAB designs for treatment)
4. Multiple Baseline Design (similar to ABA design but one receives treatment the other is baseline control)
5. Cross Sectional and Longitudinal Research

4.

Question 24

What are some advantages to Cross Sectional research?

Answer 24

Less expensive
Short time span
Mortality is minimised
Requires little cooperation between researchers and participants

Question 25

What are some disadvantages to Cross Sectional research?

Answer 25

Groups not strictly comparable
Little directional predictability
Same age doesn’t mean same maturation level
No ability to examine continuity of development

Question 26

What are some advantages to Longitudinal Research?

Answer 26

Can study development over extended period of time
Subject are their own controls
continuity between different groups
Can infer some cause and effect

Question 27

What are some disadvantages to Longitudinal Research?

Answer 27

Expense

Mortality

Question 28

What is a spurious relationship?

Answer 28

A spurious relationship or spurious correlation is a mathematical relationship in which two or more events or variables are associated but not causally related, due to either coincidence or the presence of a certain third, unseen factor.