RESEARCH METH AND DESCRIPTIVE STATISTICS
Empirical research
Empirical questions: we can measure it, we can observe it, we can use data to answer the
questions.
• NOT empirical question: questions that we can debate on, we can not observe or
measure it.
Rephrase to empirical question: Is rembrandt a better painter than rubens à do people buy
more paintings of rembrandt of more of rubens?
• Create empirical questions from non empirical questions.
Empirical questions are answered with thinking and observing.
Origin of empirical questions:
• Curiosity
• Science
• Decision-making
o Circle: problem & need analysis (causes, consequences, how big is the
problem?) à find & design options à ex ante options evaluation à choice
à implementation à ex post evaluation (effect / impact research)
§ Loneleness among elderly: are they really feeling lonely (descriptive)
à options: has someone used an option to solve this problem before?
à can we expect the option to work? What is the effect of the
option? à we chose an option à start with an option and do process
evaluation. Did we do as plannend? Did we do the interfancing as
plannend? à did the select options have the outcome we wanted?
Outcome evaluation?
§ Ex ante evaluation is research to check whether possible solutions
might actually work (so before making a decision)
§ Process evaluation is checking whether the solution was implemented
(and why (not)
§ Outcome evaluation is about whether the solution actually worked.
Two types of empirical questions
• Descriptive: to describe (how many students are on UT?) Measure it and give a level.
o What is the stress level?
o What are the sleeping patterns?
• Explanatory (or causal): about cause and defect (why is a student on the UT?)
Something influences another thing. Explain the effect of the cause.
o Do increased levels of stress lead to decreased levels of health?
o Is the drinking behavior of students influenced by answering questions about
excessive drinking?
Conformation bias: we only see what we want to see à why we have procedure in
empirical research. With a good research design you can make sure it doesn’t happen.
• Empirical research question à clear thinking and observing à knowledge question
o Procedure empirical research
1. Thinking about the problem (theory)
2. Planning the research (research design)
3. Observing (data collection)
1
, 4. Analyzing (data analysis)
Wheel of science: question à theory à research design à data collection à data analysis
à answers/knowledge à question à etc.
• Did we find what we are looking for?
• Sometimes an answer leads to another question.
o Not a logistical sequence of steps.
Induction and deduction
Induction:
• From a specific situation to general situation.
• Swans are white. A lot of swans are white. Are all swans are white?
Deduction:
• Swans are white. Let’s go to 3 parks to see the swans. You can not focus on the
population. You focus a small group.
• Specific situation.
Example
(Empirical) question: are diverse teams more successful?
Theory (think first): what is success? What is diversity? Why would diversity (age, sexuality)
cause success? What diversity are we looking at?
Design (planning): think about the design of the research. What are we going to research?
Data collection and analysis (spss) and visualization (graphs and tables).
Answer/conclusion: diverse teams (conceptualized as …) are more (less) creative (meaning
…) than homogeneous teams.
Clear empirical questions
Types of questions:
• Normative: what should be the case?
• Conceptual: what is the meaning of the case? What is the meaning of the word?
o Start of research
• Empirical: what is the case? We can observe/measure this.
o Explanatory: cause/effect.
§ One variable is cause and one is effect.
§ Dependent and independent variable.
o Descriptive: describe/explanation
Research question contains:
• Units: how or what are we comparing with each other.
• Variables: what we can the units measure with. We compare the units (people bijv.)
on this.
o Dependent variables = effect
o Independent variables = cause
§ Values / attributes: all variables can me measured.
• Setting: narrow the research down. Time and place.
o Countries à Europe.
Example: does increasing economic growth reduce the fertility rates in European countries?
• Explanatory question: economic growth influences the fertility rates
• Independent (cause) economic growth à dependent (effect) fertility rates
2
, • Units: European countries
• Setting: no setting. European belongs to countries, so European countries is the unit.
You can add setting (year).
Research question Normative, If empirical.
conceptual or Descriptive or
empirical? explanatory?
Are free markets better for society than Normative
regulated markets?
Should the government put more effort Normative
into stimulating Dutch children to play
outside?
How many minutes per day do children Empirical Descriptive
aged 6 to 12 in Germany in 2017 play
outside?
Which factors explain differences in the Empirical Explanatory
extent to which children aged 6 to 12 in
Germany in 2017 play outside?
What is meant by ‘playing outside’? Conceptual
How many German children play outside Empirical Descriptive
for more than 30 minutes per day (on
average)?
Should euthanasia be outlawed? Normative
What is the effect of playing computer Empirical Explanatory
games on the study results of UT students
in 2017?
Research question Variable(s)? Unit(s) of Setting
analysis?
What are the effects of three - Drinking behaviour People Germany
alcoholic drinks on driving a - Quality of driving between the
bicycle among people between behaviour. age 65 and 75
the age 65 and 75 in Germany?
Does the smell of chocolate, - Smell of chocolate Women During/after
during a cognitive test, make Attributes: yes a cognitive
women more craving for (smell chocolate) or test
chocolate after the test? no (not smell
chocolate)
- Craving chocolate
Do people shopping spend more - Hearing music People (how -
money when they hear music - Amount of are shopping)
during shopping? spending
3
, Do Dutch high school students in - Type of reading Dutch high 2017
2017 score better on a task used school students
computerized summative reading - Scoring on the task
task or on a paper-and-pencil
summative reading task?
Does medicine A have less sides- Getting medicine a Boys between
effect than medicine B for boys or b 10 and 20 years
between 10 and 20-year old with Side effects old with severe
severe acne and do the side- Sitting in the sun acne
effects increase when sitting in
the sun?
Medicine à side effects
^ ^ (effects on medicine and side
effects)
Sitting in the sun
Variable What are the attributes/values?
Gender Female, male
Income From – infinity to + intinity (well)
Satisfaction with democracy An attitude, the number of attributes very much
depends on the exact scale used. A 5 or 10 point scale.
Religion Christianity, Hinduism, Buddhism, other.
Car brands Toyota, Volkswagen, Nissan, ford, other.
Current Relationship status Single, living together, married, other.
Data
Units of analysis: the units we are interested in.
• Ask questions about and formulate hypotheses about units of a.
Units of observation: the units we get the information from.
• Collect data about units of o.
• Example: heterogeneity of teams -> units of analysis are the teams. Units of
observation are the individuals.
• Exercise:
Example Unit of analysis Unit of
observation
A researcher wants to investigate the general Individuals living Individuals living
attitude towards immigrants in Enschede. Therefore, in in
she randomly selects citizens of Enschede to Enschede/citizens Enschede/citizens
participate in a survey.
A researcher wants to study if certain company Companies HR managers
characteristics can explain why some companies
have a higher percentage of employees on sick leave
4
Empirical research
Empirical questions: we can measure it, we can observe it, we can use data to answer the
questions.
• NOT empirical question: questions that we can debate on, we can not observe or
measure it.
Rephrase to empirical question: Is rembrandt a better painter than rubens à do people buy
more paintings of rembrandt of more of rubens?
• Create empirical questions from non empirical questions.
Empirical questions are answered with thinking and observing.
Origin of empirical questions:
• Curiosity
• Science
• Decision-making
o Circle: problem & need analysis (causes, consequences, how big is the
problem?) à find & design options à ex ante options evaluation à choice
à implementation à ex post evaluation (effect / impact research)
§ Loneleness among elderly: are they really feeling lonely (descriptive)
à options: has someone used an option to solve this problem before?
à can we expect the option to work? What is the effect of the
option? à we chose an option à start with an option and do process
evaluation. Did we do as plannend? Did we do the interfancing as
plannend? à did the select options have the outcome we wanted?
Outcome evaluation?
§ Ex ante evaluation is research to check whether possible solutions
might actually work (so before making a decision)
§ Process evaluation is checking whether the solution was implemented
(and why (not)
§ Outcome evaluation is about whether the solution actually worked.
Two types of empirical questions
• Descriptive: to describe (how many students are on UT?) Measure it and give a level.
o What is the stress level?
o What are the sleeping patterns?
• Explanatory (or causal): about cause and defect (why is a student on the UT?)
Something influences another thing. Explain the effect of the cause.
o Do increased levels of stress lead to decreased levels of health?
o Is the drinking behavior of students influenced by answering questions about
excessive drinking?
Conformation bias: we only see what we want to see à why we have procedure in
empirical research. With a good research design you can make sure it doesn’t happen.
• Empirical research question à clear thinking and observing à knowledge question
o Procedure empirical research
1. Thinking about the problem (theory)
2. Planning the research (research design)
3. Observing (data collection)
1
, 4. Analyzing (data analysis)
Wheel of science: question à theory à research design à data collection à data analysis
à answers/knowledge à question à etc.
• Did we find what we are looking for?
• Sometimes an answer leads to another question.
o Not a logistical sequence of steps.
Induction and deduction
Induction:
• From a specific situation to general situation.
• Swans are white. A lot of swans are white. Are all swans are white?
Deduction:
• Swans are white. Let’s go to 3 parks to see the swans. You can not focus on the
population. You focus a small group.
• Specific situation.
Example
(Empirical) question: are diverse teams more successful?
Theory (think first): what is success? What is diversity? Why would diversity (age, sexuality)
cause success? What diversity are we looking at?
Design (planning): think about the design of the research. What are we going to research?
Data collection and analysis (spss) and visualization (graphs and tables).
Answer/conclusion: diverse teams (conceptualized as …) are more (less) creative (meaning
…) than homogeneous teams.
Clear empirical questions
Types of questions:
• Normative: what should be the case?
• Conceptual: what is the meaning of the case? What is the meaning of the word?
o Start of research
• Empirical: what is the case? We can observe/measure this.
o Explanatory: cause/effect.
§ One variable is cause and one is effect.
§ Dependent and independent variable.
o Descriptive: describe/explanation
Research question contains:
• Units: how or what are we comparing with each other.
• Variables: what we can the units measure with. We compare the units (people bijv.)
on this.
o Dependent variables = effect
o Independent variables = cause
§ Values / attributes: all variables can me measured.
• Setting: narrow the research down. Time and place.
o Countries à Europe.
Example: does increasing economic growth reduce the fertility rates in European countries?
• Explanatory question: economic growth influences the fertility rates
• Independent (cause) economic growth à dependent (effect) fertility rates
2
, • Units: European countries
• Setting: no setting. European belongs to countries, so European countries is the unit.
You can add setting (year).
Research question Normative, If empirical.
conceptual or Descriptive or
empirical? explanatory?
Are free markets better for society than Normative
regulated markets?
Should the government put more effort Normative
into stimulating Dutch children to play
outside?
How many minutes per day do children Empirical Descriptive
aged 6 to 12 in Germany in 2017 play
outside?
Which factors explain differences in the Empirical Explanatory
extent to which children aged 6 to 12 in
Germany in 2017 play outside?
What is meant by ‘playing outside’? Conceptual
How many German children play outside Empirical Descriptive
for more than 30 minutes per day (on
average)?
Should euthanasia be outlawed? Normative
What is the effect of playing computer Empirical Explanatory
games on the study results of UT students
in 2017?
Research question Variable(s)? Unit(s) of Setting
analysis?
What are the effects of three - Drinking behaviour People Germany
alcoholic drinks on driving a - Quality of driving between the
bicycle among people between behaviour. age 65 and 75
the age 65 and 75 in Germany?
Does the smell of chocolate, - Smell of chocolate Women During/after
during a cognitive test, make Attributes: yes a cognitive
women more craving for (smell chocolate) or test
chocolate after the test? no (not smell
chocolate)
- Craving chocolate
Do people shopping spend more - Hearing music People (how -
money when they hear music - Amount of are shopping)
during shopping? spending
3
, Do Dutch high school students in - Type of reading Dutch high 2017
2017 score better on a task used school students
computerized summative reading - Scoring on the task
task or on a paper-and-pencil
summative reading task?
Does medicine A have less sides- Getting medicine a Boys between
effect than medicine B for boys or b 10 and 20 years
between 10 and 20-year old with Side effects old with severe
severe acne and do the side- Sitting in the sun acne
effects increase when sitting in
the sun?
Medicine à side effects
^ ^ (effects on medicine and side
effects)
Sitting in the sun
Variable What are the attributes/values?
Gender Female, male
Income From – infinity to + intinity (well)
Satisfaction with democracy An attitude, the number of attributes very much
depends on the exact scale used. A 5 or 10 point scale.
Religion Christianity, Hinduism, Buddhism, other.
Car brands Toyota, Volkswagen, Nissan, ford, other.
Current Relationship status Single, living together, married, other.
Data
Units of analysis: the units we are interested in.
• Ask questions about and formulate hypotheses about units of a.
Units of observation: the units we get the information from.
• Collect data about units of o.
• Example: heterogeneity of teams -> units of analysis are the teams. Units of
observation are the individuals.
• Exercise:
Example Unit of analysis Unit of
observation
A researcher wants to investigate the general Individuals living Individuals living
attitude towards immigrants in Enschede. Therefore, in in
she randomly selects citizens of Enschede to Enschede/citizens Enschede/citizens
participate in a survey.
A researcher wants to study if certain company Companies HR managers
characteristics can explain why some companies
have a higher percentage of employees on sick leave
4
