Midterm | Tilburg University | Learning statistics with Jamovi
Number (type) Levels IV Number Type DV Design Analysis
IV (if cat.) DV
0 - 1 continuous against fixed One sample T-test
value Homogeneity not an issue: 1 var
1 (categorical) 2 1 continuous between subjects Independent samples T-test
1 (categorical) 2 1 continuous within subjects Dependent samples T-test
(large indiv. (i.e. paired samples)
differences) Homogeneity not an issue: 1 group
1 (categorical) >3 1 continuous between subjects One-way ANOVA
> 2 (categorical) >2 1 continuous between subjects Factorial ANOVA
1 (categorical) >2 1 categorical between subjects Pearson Chi-square
1 (continuous) - 1 continuous - Correlation / (simple) regression
Example Hypotheses (H1) Example Null Hypotheses (H0)
One sample T-test H1: The percentage of correct judgments for
(fixed value) non-words is more than 50%
H1: The sample is, on average, older than 20
years old
Independent T-test H1: On average, boys drink more units of
(between samples) alcohol per week than girls.
Dependent T-test H1: People’s correctness judgment of real
(i.e. paired English words is better than their correctness
samples) judgment of non-words
(within subjects)
H1: After playing a violent video game, people
Before-after design show more aggressive behavior.
H1: Children conduct more mischievous acts
when they are told they are invisible compared
to when they are told they can be seen.
1
,CONDUCTING T-TESTS
Steps involved in conducting T-Tests, with considerations for data preparation, handling
homogeneity issues, and reporting results.
ANOVA | Hypotheses
Example Hypotheses (H1) Example Null Hypotheses (H0)
One-way H1: Respondents with stress about their health H0: The three sources of stress (health,
ANOVA perceive more stress than respondents who children, finances) lead to equal amounts
(between stress about their children or finances of perceived stress.
samples) (= specific = PLANNED CONTRAST)
(IV’s with H1: On average, there is a difference in the H0: On average, teachers from media,
multiple student-perceived physical attractiveness of humanities, and technical faculties are equal in
levels) teachers from media, humanities, and the student-perceived physical attractiveness.
technical faculties
(= specific = PLANNED CONTRAST)
H1: Younger people are less optimistic in life H0: Younger people are equally optimistic in
than older people life as older people.
(= not specific = POST-HOC)
2
,CONDUCTING ANOVA
Steps involved in conducting an ANOVA, with considerations for data preparation, handling
homogeneity issues, and reporting results.
3
, Week 1 | Introduction in doing research | Chapter 2 & 3
The research process
Null-hypothesis (H0): assumption there is no effect - always try to reject H0
● e.g. Woman are equally likely as men to wear a skirt or dress
● e.g. There is no relationship between age and the number of wrinkles you have
● If the absence of difference explains your data well > accept H0
Alternative hypothesis (H1): the one you want to test, why you started it
● If you can reject H0, then H1 is SUPPORTED by the data (not PROVEN)
● e.g. Women are more likely to wear a skirt or dress than men
● e.g. there is a positive relationship between age and the number of wrinkles you
have: the older people are, the more wrinkles you have
Why do we need statistics?
● Statistics offer us a means to determine the chance the null hypothesis is true,
or: how (un-)likely it is that we would observe a set of data if H0 were true.
● If it is very unlikely (<5%) = there is support for alternative hypothesis (H1)
Research designs
Experimental Example: Nicole with glasses vs.
● manipulate 1 variable
Design Nicole without glasses (IV) and
● look for cause-and-effect
job prospects (DV)
relationships
● manipulated variable (the cause) is
the independent variable
● effect is the dependent variable
Correlational
● You can NOT manipulate this associations | e.g. is depression
Design
● Measure/observe (perceived) reality associated with bad health?
● Identifying the presence and
predictor (outcome) variable |
strength of a relationship between
Does lecture attendance predict
variables
grade?
While correlations suggest connections,
= a (strong) correlation not it’s
they do NOT establish causation.
not a cause
4
Number (type) Levels IV Number Type DV Design Analysis
IV (if cat.) DV
0 - 1 continuous against fixed One sample T-test
value Homogeneity not an issue: 1 var
1 (categorical) 2 1 continuous between subjects Independent samples T-test
1 (categorical) 2 1 continuous within subjects Dependent samples T-test
(large indiv. (i.e. paired samples)
differences) Homogeneity not an issue: 1 group
1 (categorical) >3 1 continuous between subjects One-way ANOVA
> 2 (categorical) >2 1 continuous between subjects Factorial ANOVA
1 (categorical) >2 1 categorical between subjects Pearson Chi-square
1 (continuous) - 1 continuous - Correlation / (simple) regression
Example Hypotheses (H1) Example Null Hypotheses (H0)
One sample T-test H1: The percentage of correct judgments for
(fixed value) non-words is more than 50%
H1: The sample is, on average, older than 20
years old
Independent T-test H1: On average, boys drink more units of
(between samples) alcohol per week than girls.
Dependent T-test H1: People’s correctness judgment of real
(i.e. paired English words is better than their correctness
samples) judgment of non-words
(within subjects)
H1: After playing a violent video game, people
Before-after design show more aggressive behavior.
H1: Children conduct more mischievous acts
when they are told they are invisible compared
to when they are told they can be seen.
1
,CONDUCTING T-TESTS
Steps involved in conducting T-Tests, with considerations for data preparation, handling
homogeneity issues, and reporting results.
ANOVA | Hypotheses
Example Hypotheses (H1) Example Null Hypotheses (H0)
One-way H1: Respondents with stress about their health H0: The three sources of stress (health,
ANOVA perceive more stress than respondents who children, finances) lead to equal amounts
(between stress about their children or finances of perceived stress.
samples) (= specific = PLANNED CONTRAST)
(IV’s with H1: On average, there is a difference in the H0: On average, teachers from media,
multiple student-perceived physical attractiveness of humanities, and technical faculties are equal in
levels) teachers from media, humanities, and the student-perceived physical attractiveness.
technical faculties
(= specific = PLANNED CONTRAST)
H1: Younger people are less optimistic in life H0: Younger people are equally optimistic in
than older people life as older people.
(= not specific = POST-HOC)
2
,CONDUCTING ANOVA
Steps involved in conducting an ANOVA, with considerations for data preparation, handling
homogeneity issues, and reporting results.
3
, Week 1 | Introduction in doing research | Chapter 2 & 3
The research process
Null-hypothesis (H0): assumption there is no effect - always try to reject H0
● e.g. Woman are equally likely as men to wear a skirt or dress
● e.g. There is no relationship between age and the number of wrinkles you have
● If the absence of difference explains your data well > accept H0
Alternative hypothesis (H1): the one you want to test, why you started it
● If you can reject H0, then H1 is SUPPORTED by the data (not PROVEN)
● e.g. Women are more likely to wear a skirt or dress than men
● e.g. there is a positive relationship between age and the number of wrinkles you
have: the older people are, the more wrinkles you have
Why do we need statistics?
● Statistics offer us a means to determine the chance the null hypothesis is true,
or: how (un-)likely it is that we would observe a set of data if H0 were true.
● If it is very unlikely (<5%) = there is support for alternative hypothesis (H1)
Research designs
Experimental Example: Nicole with glasses vs.
● manipulate 1 variable
Design Nicole without glasses (IV) and
● look for cause-and-effect
job prospects (DV)
relationships
● manipulated variable (the cause) is
the independent variable
● effect is the dependent variable
Correlational
● You can NOT manipulate this associations | e.g. is depression
Design
● Measure/observe (perceived) reality associated with bad health?
● Identifying the presence and
predictor (outcome) variable |
strength of a relationship between
Does lecture attendance predict
variables
grade?
While correlations suggest connections,
= a (strong) correlation not it’s
they do NOT establish causation.
not a cause
4
