(PM) BRT - Statistical Methods Flashcards
Nominal (IV) - Nominal (DV)
Chi-square test
Ordinal (IV) - Nominal (DV)
Chi-square test
Nominal (IV) - Ordinal (DV)
Chi-square test
Ordinal (IV) - Ordinal (DV)
Chi-square test
Nominal (IV) - Interval (DV)
T-test, ANOVA
Ordinal (IV) - Interval (DV)
T-test, ANOVA
Nominal (IV) - Ratio (DV)
T-test, ANOVA
Ordinal (IV) - Ratio (DV)
T-test, ANOVA
Interval (IV) - Nominal (DV)
Logit analysis
Ratio (IV) - Nominal (DV)
Logit analysis
Interval (IV) - Ordinal (DV)
Logit analysis
Ratio (IV) - Ordinal (DV)
Logit analysis
Interval (IV) - Interval (DV)
Regression analysis
Ratio (IV) - Interval (DV)
Regression analysis
Interval (IV) - Ratio (DV)
Regression analysis
Ratio (IV) - Ratio (DV)
Regression analysis
When I measure the IV ‘class attendance’ with 2 levels (low attendance vs. high attendance ) and want to test its relationship with the ratio-scaled DV ‘grades’, I should carry out …
T-test (or One-Way ANOVA in case of factorial experiment)
When I measure the IV ‘class attendance’ and the IV ‘Assignment result’ and want to test its relationship with the ratio-scaled DV ‘grades’, I should carry out …
ANOVA (in case of factorial experiment, Two-Way)
When I measure IV “University status” (with levels; Failed Pre-Master, Passed Pre-Master and Graduated) and want to test its relationship with ‘Income’, I should carry out …
ANOVA (in case of factorial experiment, One-Way)
Stratified Sampling
Divide in homogenous groups, then apply SRS
Cluster Sampling
Divide in heterogenous groups, then select a number of groups
A researcher wants to study the effect of packaging shape (which can be either squared vs round) and packaging height (where height can take on any value between 0 and 100 centimetres) on perceived healthiness (7-point Likert scale).
Which test should be used?
Regression Analysis
History effect
Events/factors outside the experiment that have an impact on the DV
Maturation Effect
Biological / Psychological changes over time
Testing Effect
Prior testing affecting the DV
Instrumentation Effect
Observed effect being due to a change in measurement
Selection Bias Effect
Incorrect selection of respondents
Mortality effect!
Drop-out of respondents during experiment
Symbolic Notation: CG
Control group
Symbolic Notation: EG
Experimental group
Symbolic Notation: X(i)
Exposure to IV treatment i
Symbolic Notation: ‘ ‘
No IV treatment.
Symbolic Notation: O(i)
Observation of DV i
Symbolic Notation: R
Random allocation of respondents to treatment
X O1 O2
No random allocation. Participants are all given treatment X. DV 1 and 2 are measured.
R X1 O1
Random allocation to treatment 1. DV 1 is measured
R X2 O1
Random allocation to treatment 2. DV 1 is measured
R O1
Random allocation, but no treatment. DV 1 is measured
One-shot Case Study
Cause-effect study. Notation: X O. Remark: no comparisons possible
One-group Pretest-posttest
Investigate influence of a variable. Notation: O X O. Remark: no control for extraneous variables
Static group
Comparing a treatment with a control group. Notation:
EG: X O1
CG: O2
Remark: there can be a difference in the EG and CG
Pretest-posttest control group
Controls for confounds (variables that threaten internal validity). Notation:
EG: R 01 X O2
CG: R O3 O4
Posttest-only control group
When pre-tests are nog possible. Notation:
EG: R X O1
CG: R O2
More simple than pretest posttest control group. Mortality effects possible
Solomon Four Group
Minimize effects of pre-testing. Notation:
EG1: R O1 X O2
CG1: R O3 O4
EG2: R X O5
CG2: R O6
Strongest form of experimental control. Takes testing effects into account.
Time series
When one group is available, and little control over manipulation and participants.
Notation: O1 O2 O3 X O4 O5
Sensitive for history and re-test effects
Multiple time series
Time series with control group Notation: EG: O1 O2 O3 X O4 O5 CG: O7 O8 O9 Different tests are possible (pre vs post, EG vs CG).
Which test should be used for the following full factorial experiment?
One categorical IV with two levels, one continuous DV
T-Test or One-Way ANOVA
Which test should be used for the following full factorial experiment?
One categorical IV with four levels, one continuous DV
One-Way ANOVA
Which test should be used for the following full factorial experiment?
Two categorical IVs with both two levels, one continuous DV
Two-Way ANOVA
Which test should be used for the following full factorial experiment? One IV (Education) with three levels (MBO, HBO, WO), one manipulated moderator continuous DV
Two-Way ANOVA
What is a heterogeneous treatment effect?
Same treatment may affect different individuals differently
Deductive research Process
- Define the business problem
- Formulate the problem statement
- Develop theoretical background
- Choose a research design
- Collect data
- Analyze data
- Write-up
Inductive Research Process
- Define the business problem
- Formulate the problem statement
- Provide a conceptual background
- Choose a research design
- Collect data
- Analyze data
- Develop theory
In exploratory research, researchers strive to …
… capture details, and rich data, obtain ‘saturation’ in insights from participants and to generate theory.
In exploratory research, researchers strive to …
… capture details, and rich data, obtain ‘saturation’ in insights from participants and to generate theory.
What forms for international research exists?
Sequential-, independent- and simultaneous multi-country research.
Guaranteeing construct equivalence (in international research) can be done using…
Drawing from country-specific literature and conducting qualitative research (to identify cultural differences)
Deductive research (types)
Causal & Correlational research
Inductive research (types)
Exploratory research
Validity
Measuring what you want to measure
Reliability
Accuracy, exactness, replicability.
Proxy
Approximation of underlying construct (within archival research validity)
Deductive research (sampling)
Probability sampling
Often used sampling technique in Inductive research
Non-probability sampling
What strategies fall under causal research?
Lab ad field experiments.
What strategies fall under correlational research?
Archival and survey research
Controlled vs uncontrolled
Lab vs Natural environment
Participant vs non-participant
Is the observator a part of the observation or not? Is the observator taking part in the activity
Structured vs Unstructured
Does the observator have a list of things they want to observe or not?
Concealed vs Non-concealed
Does the participant know they are being observed?
Multivariate research
Multiple variables in a study, and where regression is used.
Moderators…
Affect the relationships between variables, and have a direct effect on the dependent variable.
Purposiveness
The ‘Why’ of the research
Rigor
‘Strict precision’ or ‘Exactness’
Objectivity
Making conclusions based on facts, not on personal (or supervisor driven) feelings or emotions
Parsimony
Shave away unnecessary details
Replicability
If someone replicates your study, they should find the same results
Generalizability
Being able to apply the finding in a wide variety of settings
