Methods, Measurement and Statistics
Premaster Program – Fall 2022
Let’s have a closer look.
The journal reported a figure from country X’s national bureau of statistics (NBS).
Detailed information from NBS also indicated that:
80% suffered from mild burnout and were on sick leave for 2 months;
15% had regular amount of burnout and recovered within 6 months;
and the remaining 5% did not work for the full year.
Assume:
10.000.000 employees of which 1.600.000 (16%) burnouts
264 working days per year (= 22 on average per month)
Total number of working days in one year = 10.000.000 * 264 = 2.640.000.000 (a)
Total number of working days lost by sick leave due to burnout = 109.120.000 (table below)
=> 4,13% of (a) or +- 11 working days on average per employee per year
Prelude: on ‘naïve’ statistics
A journal reported that the economy of the country was under severe stress in 2021
because of the fact that approximately 16% of all employees suffered from burnout.
Seems very impressive, isn’t it?
Lessons learned?
Be prudent with just looking at one figure, you need to put it in perspective.
Don’t let emotions run ahead of your logical thinking, be critical.
Be cleverer than the average journalist
1. Cornerstones of social research (chapter 2)
theory(-ies) <- propositions -> hypotheses
proposition:
example “When an individual manages a particular task well, then (s)he will perform that
task better in the presence of others than when nobody else is present.”
(= social facilitation effect SFE)
= general statement regarding a regularity in the behavior or opinion of subjects
The question is: why is this the case?
Theory = provides an explanation for a proposition or set of propositions ≠ speculation! ex.
Alternative theories for the SFE proposition: biological versus psychological ‘explanation’
- ‘biological’: the presence of others activates physiological triggers
- ‘psychological’: people perform better when they believe they are being
watched/evaluated
= nature versus nurture debate in many social and behavioral sciences
1
,Next question: how can we research that?
= applying it in a concrete situation
e.g.: athletics
-> hypotheses = athletes will perform better:
(a) the more spectators there are;
(b) the more journalists there are that will comment on their performances;
(c) when there is direct coverage of the event by the media (television,…)
Distinguishing proposition from hypothesis – a class example
A well-known ‘theorem’ in the social sciences states that (Thomas theorem)
“If men define situations as real, they are real in their consequences.”
Assignment: define a proposition and a hypothesis that can be matched with this theorem
in the context of ‘superstition’
Proposition: When someone strongly beliefs in the existence of invisible creatures having
certain powers, the alleged powers assigned to such creatures become real in their
consequences.
Hypothesis: The more convinced children are that invisible monsters live under their bed
waiting to haunt them in the dark, the more likely these children will experience nightmares
that prevent them from sleeping in dark bedrooms.
Curious how this would translate into a multiple choice exam question?
Example multiple choice question:
Evaluate the following statement:
“The more convinced children are that monsters live under their bed waiting to haunt them
when it is dark, the more likely these children will experience nightmares that prevent them
from sleeping in dark bedrooms.” is an example of a hypothesis rather than a proposition.
TRUE or FALSE? => TRUE
2
,2. Science as a process: induction – deduction
(chapter 2)
theories deduction
empirical
hypotheses
generalisations
induction observations
Inductively induced research project – example
Johan Denollet’s “D-personality”
- He was working as a psychologist at the cardiac rehabilitation unit of the Antwerp
University Hospital when he observed that different patients had different chances
of rehabilitation that seems to be linked to personality characteristics.
- He then developed a method that allowed measuring two aspects of personality
(‘negative affectivity’ (NA) and ‘social inhibition’(SI)) and found that when patients
combined both of them these patients had less chance of rehabilitation.
- He then developed his theory on the D-type personality (D = distress = combining NA
+ SI) arguing that this broad and stable personality trait is a clear risk factor of
cardiac events in the future.
Deductively induced research project – example. Peter M. Blau and Otis D. Duncan:
- Developed the status attainment model in the 1960’s to explain social mobility
patterns in societies.
- The overarching research question was: what attributes facilitate the movement of
individuals into occupations with particular (desirable) social status?
- They argued that occupational outcomes are shaped by family of origin in a direct
and indirect way. A typical model is as follows:
Reading 2x2 tables
Example: Gender bias in hiring employees?
We focus on “not hired” calculating % not hired
F = 125/180 = 69%; M = 344/550 = 63%
Difference: 6% (69% - 63%) more female applicants not hired than male applicants
3
, Concepts, variables and hypotheses (Chapter 3)
Concepts (or constructs) = general/abstract description of a social phenomena
e.g. ethnocentrism
Variable = empirical manifestation of a concept
e.g. a scale that measures ethnocentrism
Hypotheses = an expected relationship between 2 or more variables that can be
researched/tested
e.g. women are on average less ethnocentric than men
Defining hypotheses – requirements
= expected relationship between 2 or more variables
Wrong: women are clever
Correct: women are cleverer than men
! Hence: do not confuse ‘characteristic’ (e.g. women) with ‘variable’ (e.g. gender)
- wording requirement: wording should reflect measurement level of the variables
involved
Gender = nominal (categories) and Intelligence = metric hypothesis regarding
difference in mean scores of intelligence for different categories of gender
4. Types of hypotheses (chapter 3 + pages 463-
470)
4.1 Bivariate hypothesis: expected relationship between two variables (= total effect)
diagram:
X Y
legend:
X = independent variable (‘cause’)
Y = dependent variables (‘outcome’)
= direction of effect (from independent on dependent)
Examples:
“The higher the emotional intelligence of a person, the higher the amount of money a
person gives to good causes.”
Wording reflects metric measurement (scale)
“The higher the emotional intelligence of a person the higher the likelihood of becoming
depressed.”
Dependent = categorical (non metric);
independent = metric
“Women have on average a higher level of emotional intelligence than men.”
Independent = categorical (non-metric); dependent =
metric
- ! Distinction between metric versus non-metric refers to ‘measurement level’ of a
variable (see later).
- ! Making the distinction is vital to statistics.
- ! Formulation of hypothesis should be consistent with this distinction.
- ! Applies to all types of hypotheses.
4
Premaster Program – Fall 2022
Let’s have a closer look.
The journal reported a figure from country X’s national bureau of statistics (NBS).
Detailed information from NBS also indicated that:
80% suffered from mild burnout and were on sick leave for 2 months;
15% had regular amount of burnout and recovered within 6 months;
and the remaining 5% did not work for the full year.
Assume:
10.000.000 employees of which 1.600.000 (16%) burnouts
264 working days per year (= 22 on average per month)
Total number of working days in one year = 10.000.000 * 264 = 2.640.000.000 (a)
Total number of working days lost by sick leave due to burnout = 109.120.000 (table below)
=> 4,13% of (a) or +- 11 working days on average per employee per year
Prelude: on ‘naïve’ statistics
A journal reported that the economy of the country was under severe stress in 2021
because of the fact that approximately 16% of all employees suffered from burnout.
Seems very impressive, isn’t it?
Lessons learned?
Be prudent with just looking at one figure, you need to put it in perspective.
Don’t let emotions run ahead of your logical thinking, be critical.
Be cleverer than the average journalist
1. Cornerstones of social research (chapter 2)
theory(-ies) <- propositions -> hypotheses
proposition:
example “When an individual manages a particular task well, then (s)he will perform that
task better in the presence of others than when nobody else is present.”
(= social facilitation effect SFE)
= general statement regarding a regularity in the behavior or opinion of subjects
The question is: why is this the case?
Theory = provides an explanation for a proposition or set of propositions ≠ speculation! ex.
Alternative theories for the SFE proposition: biological versus psychological ‘explanation’
- ‘biological’: the presence of others activates physiological triggers
- ‘psychological’: people perform better when they believe they are being
watched/evaluated
= nature versus nurture debate in many social and behavioral sciences
1
,Next question: how can we research that?
= applying it in a concrete situation
e.g.: athletics
-> hypotheses = athletes will perform better:
(a) the more spectators there are;
(b) the more journalists there are that will comment on their performances;
(c) when there is direct coverage of the event by the media (television,…)
Distinguishing proposition from hypothesis – a class example
A well-known ‘theorem’ in the social sciences states that (Thomas theorem)
“If men define situations as real, they are real in their consequences.”
Assignment: define a proposition and a hypothesis that can be matched with this theorem
in the context of ‘superstition’
Proposition: When someone strongly beliefs in the existence of invisible creatures having
certain powers, the alleged powers assigned to such creatures become real in their
consequences.
Hypothesis: The more convinced children are that invisible monsters live under their bed
waiting to haunt them in the dark, the more likely these children will experience nightmares
that prevent them from sleeping in dark bedrooms.
Curious how this would translate into a multiple choice exam question?
Example multiple choice question:
Evaluate the following statement:
“The more convinced children are that monsters live under their bed waiting to haunt them
when it is dark, the more likely these children will experience nightmares that prevent them
from sleeping in dark bedrooms.” is an example of a hypothesis rather than a proposition.
TRUE or FALSE? => TRUE
2
,2. Science as a process: induction – deduction
(chapter 2)
theories deduction
empirical
hypotheses
generalisations
induction observations
Inductively induced research project – example
Johan Denollet’s “D-personality”
- He was working as a psychologist at the cardiac rehabilitation unit of the Antwerp
University Hospital when he observed that different patients had different chances
of rehabilitation that seems to be linked to personality characteristics.
- He then developed a method that allowed measuring two aspects of personality
(‘negative affectivity’ (NA) and ‘social inhibition’(SI)) and found that when patients
combined both of them these patients had less chance of rehabilitation.
- He then developed his theory on the D-type personality (D = distress = combining NA
+ SI) arguing that this broad and stable personality trait is a clear risk factor of
cardiac events in the future.
Deductively induced research project – example. Peter M. Blau and Otis D. Duncan:
- Developed the status attainment model in the 1960’s to explain social mobility
patterns in societies.
- The overarching research question was: what attributes facilitate the movement of
individuals into occupations with particular (desirable) social status?
- They argued that occupational outcomes are shaped by family of origin in a direct
and indirect way. A typical model is as follows:
Reading 2x2 tables
Example: Gender bias in hiring employees?
We focus on “not hired” calculating % not hired
F = 125/180 = 69%; M = 344/550 = 63%
Difference: 6% (69% - 63%) more female applicants not hired than male applicants
3
, Concepts, variables and hypotheses (Chapter 3)
Concepts (or constructs) = general/abstract description of a social phenomena
e.g. ethnocentrism
Variable = empirical manifestation of a concept
e.g. a scale that measures ethnocentrism
Hypotheses = an expected relationship between 2 or more variables that can be
researched/tested
e.g. women are on average less ethnocentric than men
Defining hypotheses – requirements
= expected relationship between 2 or more variables
Wrong: women are clever
Correct: women are cleverer than men
! Hence: do not confuse ‘characteristic’ (e.g. women) with ‘variable’ (e.g. gender)
- wording requirement: wording should reflect measurement level of the variables
involved
Gender = nominal (categories) and Intelligence = metric hypothesis regarding
difference in mean scores of intelligence for different categories of gender
4. Types of hypotheses (chapter 3 + pages 463-
470)
4.1 Bivariate hypothesis: expected relationship between two variables (= total effect)
diagram:
X Y
legend:
X = independent variable (‘cause’)
Y = dependent variables (‘outcome’)
= direction of effect (from independent on dependent)
Examples:
“The higher the emotional intelligence of a person, the higher the amount of money a
person gives to good causes.”
Wording reflects metric measurement (scale)
“The higher the emotional intelligence of a person the higher the likelihood of becoming
depressed.”
Dependent = categorical (non metric);
independent = metric
“Women have on average a higher level of emotional intelligence than men.”
Independent = categorical (non-metric); dependent =
metric
- ! Distinction between metric versus non-metric refers to ‘measurement level’ of a
variable (see later).
- ! Making the distinction is vital to statistics.
- ! Formulation of hypothesis should be consistent with this distinction.
- ! Applies to all types of hypotheses.
4
