STATISTICS 3 2023
VRIJE UNIVERSITEIT - BACHELOR PSYCHOLOGY
WEEK 1
● LECTURE 1:Recapitulation
WEEK 2:
● LECTURE 2:ANOVA / Regression
WEEK 3:
● LECTURE 3:Factorial ANOVA
WEEK 4:
● LECTURE 4:ANCOVA
WEEK 5:
● LECTURE 4:Mediation / Moderation
WEEK 6:
● LECTURE 6:MANOVA/ Repeated Measures
,.. Lectures 1: RECAPITULATION WEEK 1
DEFINITION STATISTICS
● “Statistics is the science of collecting, organizing and interpreting numerical facts, which we call data.”
● Importantmatters for the application of statistics(“Applied Statistics”):
1. Selecting a samplefrom a population
a. Random vs non-random selection
2. Deciding whether a sample isrepresentative
a. You want to generalize your findings to the population, not just this sample
3. Deciding betweenDescriptiveorInferential Statistics
4. Measurement levels (NOIR)andtypes of variables (categorical/quantitative)
5. Selecting the correctstatistical analysis
6. Experimentalversusnon-experimentalresearch design
● A&F: Statistics consists of abody of methods for obtaining and analyzing data, to:
1. Design [research studies that]
2. Describe [the data to]
3. Make inferences based on these data.
Two types:
Descriptive Statistics Inferential Statistics
escriptive statistics summarize sample or Inferential statistics make predictions about population
D
population data with numbers, tables, and graphs parameters, based on a (random) sample of data
(e.g., box plots) (Estimate population parameter based upon the sample statistic)
DATA, POPULATION, SAMPLE, RELIABILITY, VALIDITY
Doing research by means ofdata: observation of characteristics
Population T hetotal set of participants, relevant for the E .g.Population parameter: average hour of self
research question study per week ofallstudents.
Sample a subset of the populationabout who the E .g.Sample statistic: average hour of self studyper
data is collected week of a randomly selectedsample of800
students
Good data is necessary to answer the research question:
● Reliability(Precision => consistency!)
● Validity (Bias)
,DESCRIPTIVE STATISTICS
Variables, measurement levels and range
Variable:measurescharacteristicsthat candifferbetween subjects
● Types:behavior-, stimulus-, subject-, physiological variables(eg: reaction time, height, level of aggression)
● Distribution:Summary of one variable
● Association:Summary of Multiple Variables
Measurement scales(NOIR):
No
minal NORDEREDcategories, no ordering, no C
U ountries
Categorical meaningful distances, no absolute zero Eye color
ategorical /
C Biological sex
Qualitative
Or dinal RDEREDcategories, ordering, no
O ancing Skills categorized as “beginners”,
D
Categorial meaningful distances, no absolute zero “intermediate”, “advanced”
point Likert scale from very satisfied to
5
very dissatisfied
disagree/neutral/agree
Interval c ategories, ordering, meaningful T emperature in Celsius (because
Continuous distances (equal distances between Temperature of 0 does not mean ab
measures), no absolute zero absence of temperature since there is still
minus temperatures)
uantitative/
Q
Numerical = equal distance between consecutive
( „Parametric values
Methods“)
Ratio c ategories, ordering, meaningful istances between 2 cities; Number of
D
Continuous distances, an absolute zero Years a person has smoked (0 years is
absence); Age
= equal distance andtrue zero point =( Kelvin - Temperature
absence of the characteristic)
Range:
Discrete measurement unit that is indivisible (1 not 1.2) Example: brothers/sisters
Continuous infinitely divisible measurement unit Example: body height
Summarized
, Graphs of Descriptive Statistics
Bar Graph Nominal /Ordinal Barsapartfrom each other
isplaying
D
Frequencies of
categorical data
Histogram Interval / Ratio arsnext toeach other (groups are
B
isplaying
D connected!)
Frequencies of ecause can divide groups into fixes
B
quantitative data sizes such as income of 1000 - 2000
Pie Chart Categorical data isplays the share/ percentage per
D
Nominal / Ordinal category
ot suitable for many values (e.g.,
N
displaying income of 63 employees)
Not suitable for continuous data
Line Graph Interval/ Ratio sed to show “a trend” of two
U
variables at a ratio level
Box Plot isplays center (= by the median) &
D
variation of data (= box & lines)
Scatterplot Interval/ Ratio Level isplays association / relation
D
between 2 variables: positive/
negative
Difference to Line Graph:
LG: for trends /SP: for relations
S team and Quantitative Data
Leaf Plots
VRIJE UNIVERSITEIT - BACHELOR PSYCHOLOGY
WEEK 1
● LECTURE 1:Recapitulation
WEEK 2:
● LECTURE 2:ANOVA / Regression
WEEK 3:
● LECTURE 3:Factorial ANOVA
WEEK 4:
● LECTURE 4:ANCOVA
WEEK 5:
● LECTURE 4:Mediation / Moderation
WEEK 6:
● LECTURE 6:MANOVA/ Repeated Measures
,.. Lectures 1: RECAPITULATION WEEK 1
DEFINITION STATISTICS
● “Statistics is the science of collecting, organizing and interpreting numerical facts, which we call data.”
● Importantmatters for the application of statistics(“Applied Statistics”):
1. Selecting a samplefrom a population
a. Random vs non-random selection
2. Deciding whether a sample isrepresentative
a. You want to generalize your findings to the population, not just this sample
3. Deciding betweenDescriptiveorInferential Statistics
4. Measurement levels (NOIR)andtypes of variables (categorical/quantitative)
5. Selecting the correctstatistical analysis
6. Experimentalversusnon-experimentalresearch design
● A&F: Statistics consists of abody of methods for obtaining and analyzing data, to:
1. Design [research studies that]
2. Describe [the data to]
3. Make inferences based on these data.
Two types:
Descriptive Statistics Inferential Statistics
escriptive statistics summarize sample or Inferential statistics make predictions about population
D
population data with numbers, tables, and graphs parameters, based on a (random) sample of data
(e.g., box plots) (Estimate population parameter based upon the sample statistic)
DATA, POPULATION, SAMPLE, RELIABILITY, VALIDITY
Doing research by means ofdata: observation of characteristics
Population T hetotal set of participants, relevant for the E .g.Population parameter: average hour of self
research question study per week ofallstudents.
Sample a subset of the populationabout who the E .g.Sample statistic: average hour of self studyper
data is collected week of a randomly selectedsample of800
students
Good data is necessary to answer the research question:
● Reliability(Precision => consistency!)
● Validity (Bias)
,DESCRIPTIVE STATISTICS
Variables, measurement levels and range
Variable:measurescharacteristicsthat candifferbetween subjects
● Types:behavior-, stimulus-, subject-, physiological variables(eg: reaction time, height, level of aggression)
● Distribution:Summary of one variable
● Association:Summary of Multiple Variables
Measurement scales(NOIR):
No
minal NORDEREDcategories, no ordering, no C
U ountries
Categorical meaningful distances, no absolute zero Eye color
ategorical /
C Biological sex
Qualitative
Or dinal RDEREDcategories, ordering, no
O ancing Skills categorized as “beginners”,
D
Categorial meaningful distances, no absolute zero “intermediate”, “advanced”
point Likert scale from very satisfied to
5
very dissatisfied
disagree/neutral/agree
Interval c ategories, ordering, meaningful T emperature in Celsius (because
Continuous distances (equal distances between Temperature of 0 does not mean ab
measures), no absolute zero absence of temperature since there is still
minus temperatures)
uantitative/
Q
Numerical = equal distance between consecutive
( „Parametric values
Methods“)
Ratio c ategories, ordering, meaningful istances between 2 cities; Number of
D
Continuous distances, an absolute zero Years a person has smoked (0 years is
absence); Age
= equal distance andtrue zero point =( Kelvin - Temperature
absence of the characteristic)
Range:
Discrete measurement unit that is indivisible (1 not 1.2) Example: brothers/sisters
Continuous infinitely divisible measurement unit Example: body height
Summarized
, Graphs of Descriptive Statistics
Bar Graph Nominal /Ordinal Barsapartfrom each other
isplaying
D
Frequencies of
categorical data
Histogram Interval / Ratio arsnext toeach other (groups are
B
isplaying
D connected!)
Frequencies of ecause can divide groups into fixes
B
quantitative data sizes such as income of 1000 - 2000
Pie Chart Categorical data isplays the share/ percentage per
D
Nominal / Ordinal category
ot suitable for many values (e.g.,
N
displaying income of 63 employees)
Not suitable for continuous data
Line Graph Interval/ Ratio sed to show “a trend” of two
U
variables at a ratio level
Box Plot isplays center (= by the median) &
D
variation of data (= box & lines)
Scatterplot Interval/ Ratio Level isplays association / relation
D
between 2 variables: positive/
negative
Difference to Line Graph:
LG: for trends /SP: for relations
S team and Quantitative Data
Leaf Plots
