Quantitative Research Methods
Lecture 1, 06-09-2021, Course Introduction
Course introduction
- Quantitative data collection
- Quantitative data analysis (using SPSS)
- How to write up your findings in an academic/professional way
Lectures
Theoretical knowledge on quantitative methods and analysis
Discussion of the course content
Course labs
Learn how to use SPSS to analyze data
Self study & online/on campus meetings
Attendance mandatory!
Online modules
Videos
Readings
Useful links
How to go from a research topic to survey questions
Using example:
Van Craen, M., & Skogan W. G. (2015). Differences and similarities in the explanation of
ethnic minority groups’ trust in the police. European Journal of Criminology, 12(3), 300-323.
Topic: Tension between police and ethnic minorities
RQ: To what extent does social capital, perceptions of police performance and perceptions
of discrimination play a role in Polish immigrants trust in the police in Belgium?
From theory…
1. Social capital theory
2. Performance theory
3. Procedural justice theory: discrimination effect how they trust their trust in the
police
… to hypotheses
1. Social capital theory does not play a role in their trust in the police
2. Performance theory does play a role in their trust in the police
3. Procedural justice theory does not play a role in their trust in the police
From hypotheses… to variables
1. Independent variable: discrimination
2. Dependent variable: trust
Dependent variable trust; Questions/Indicators (operationalizations):
1. How much or how little trust do you have in the police?
, 2. How much or how little do you trust the Belgian police to deal with the issues that are
important for you?
3. How much or how little do you trust the police to treat people with respect?
Independent variable discrimination; Questions/Indicators (operationalizations):
1. How often within the last 12 months have you felt discriminated against?
2. How strict are the police when dealing with people of Polish descent than when
dealing with people of Belgian descent?
3. The police treat all citizens in the neighbourhood equally
Control variables (demographics)
1. Gender (0= male/1= female)
2. Age (years)
3. Level of education (0= no qualification/10= university degree)
4. Length of respondents’ residence in the city of Antwerp (years)
Survey design: The art of asking questions
Keep question s short, simple and clear
NOT: when did you start you BcSS?
BUT: in what year did you start your BcSS?
Avoid vague and abstract terms
NOT: are you a proponent of legalization of drugs?
BUT: to what extent do you agree or disagree that marihuana should be legal?
Avoid double-barreled questions
NOT: does it happen that you cannot get your work done, so you take it home with
you?
BUT: how often does it happen that you cannot get your work done? If yes, do you
take the unfinished work home?
Using examples
NOT: what do you like about your work? Think of money, flexibility etc.?
BUT: how much do you enjoy your work? What do you enjoy about your work?
(open answer)
Authority effect
NOT: to what extent do you agree with the plan on the prime-minister to enforce face
masks in shops?
BUT: to what extent do you agree with the measure on wearing face masks in shops?
Other tips:
Avoid making assumptions
Avoid abbreviations and slang
Avoid negative phrased questions and double negative questions
Use a realistic time frame in your questions
Be aware of researcher effects
Know how to ask questions
The order of the questions
, The type of responses
Standardization of the survey and data collection
Final tips
Check for existing indicators: if established measures exist, use them
Behaviors, attitudes and opinions are complex and best measured with a number
(battery) of questions?
Think about answer categories
o Avoid categorical answers of scales (e.g. age) as much as possible
o Behaviors, attitudes and opinions best answered with a Likert scale
o Avoid yes/no questions as much as possible
o Avoid multi-option responses
Limit the number of questions if possible
Make sure your survey has a coherent ‘story’
Pilot your survey
What needs to be in a survey
Step 1: Introduction
Introduce the topic and the reason this survey is done
Explain the type of questions respondents will get
o But don’t give the exact questions
Informed consent
o Make this mandatory
Contact details of researcher
Ethical issues
o Do not lie or deceive a respondent, but what if ‘innocent deception’ is needed?
See also the video on BS in the module ‘How to make a survey’
Step 2: Your questions
Present your survey questions in a logical order
That are unambiguous, clear, etc.
Make sure all questions require an answer (forced response)
Step 3: Ending
End your survey with a ‘thank you’
Offer respondents the opportunity to see the end result
Give your contract details one more time
Final information on the group paper
Choose a topic with your peers, turn your topic into a RQ and turn your RQ into
operationalizable concepts
Make your own (online group) survey
o Use Qualtrics: http://leidenuniv.eu.qualtrics.com
Have 50 respondents per group member fill in your survey
o Can be a convenience sample
Report your findings in a 1000 word paper
, o No 10% extra word limit
o Deadline: Tuesday October 5th at noon
o Hand in via Brightspace
Writing it down
Introduction:
Introduce the topic
How you went from your topic to a general RQ
Discuss theory (if relevant)
Methodology
If you selected pre-existing measures, describe why and reference them. If you made
your own questions, write down how you operationalized them (don’t forget to
reference)
Describe sampling strategy
Be as detailed as possible
Critically reflect on your survey and data collection strategies
To what extent is the survey/data collection reliable and valid?
Lecture 2, 13-09-2021, How to distribute a reliable and
valid survey
A critical eye
Science vs pseudoscience
Preconceived notions of what to believe
Searching for evidence to support a preconceived belief
Ignoring evidence that would falsify a claim or belief
Disregarding alternative explanations for a phenomenon
Accepting flimsy evidence to support an extraordinary claim
Relying heavily on anecdotal evidence
Employing very little skepticism
But is all science good?
Andrew Wakefield:
- In a study of 12 children Wakefield claimed that the measles-mumps-rubella vaccine
caused autism.
- This study still has its grips on society today
- BUT 12 years later his work has been retracted, because it was a sample size he only
looked at 12 children
o Very small sample size = 12
o Cause Effect, control group; did not have that
o Falsified outcomes
Always remain critical, not all research is quite good.
- Drinking wine before bed could help you lose weight.
Lecture 1, 06-09-2021, Course Introduction
Course introduction
- Quantitative data collection
- Quantitative data analysis (using SPSS)
- How to write up your findings in an academic/professional way
Lectures
Theoretical knowledge on quantitative methods and analysis
Discussion of the course content
Course labs
Learn how to use SPSS to analyze data
Self study & online/on campus meetings
Attendance mandatory!
Online modules
Videos
Readings
Useful links
How to go from a research topic to survey questions
Using example:
Van Craen, M., & Skogan W. G. (2015). Differences and similarities in the explanation of
ethnic minority groups’ trust in the police. European Journal of Criminology, 12(3), 300-323.
Topic: Tension between police and ethnic minorities
RQ: To what extent does social capital, perceptions of police performance and perceptions
of discrimination play a role in Polish immigrants trust in the police in Belgium?
From theory…
1. Social capital theory
2. Performance theory
3. Procedural justice theory: discrimination effect how they trust their trust in the
police
… to hypotheses
1. Social capital theory does not play a role in their trust in the police
2. Performance theory does play a role in their trust in the police
3. Procedural justice theory does not play a role in their trust in the police
From hypotheses… to variables
1. Independent variable: discrimination
2. Dependent variable: trust
Dependent variable trust; Questions/Indicators (operationalizations):
1. How much or how little trust do you have in the police?
, 2. How much or how little do you trust the Belgian police to deal with the issues that are
important for you?
3. How much or how little do you trust the police to treat people with respect?
Independent variable discrimination; Questions/Indicators (operationalizations):
1. How often within the last 12 months have you felt discriminated against?
2. How strict are the police when dealing with people of Polish descent than when
dealing with people of Belgian descent?
3. The police treat all citizens in the neighbourhood equally
Control variables (demographics)
1. Gender (0= male/1= female)
2. Age (years)
3. Level of education (0= no qualification/10= university degree)
4. Length of respondents’ residence in the city of Antwerp (years)
Survey design: The art of asking questions
Keep question s short, simple and clear
NOT: when did you start you BcSS?
BUT: in what year did you start your BcSS?
Avoid vague and abstract terms
NOT: are you a proponent of legalization of drugs?
BUT: to what extent do you agree or disagree that marihuana should be legal?
Avoid double-barreled questions
NOT: does it happen that you cannot get your work done, so you take it home with
you?
BUT: how often does it happen that you cannot get your work done? If yes, do you
take the unfinished work home?
Using examples
NOT: what do you like about your work? Think of money, flexibility etc.?
BUT: how much do you enjoy your work? What do you enjoy about your work?
(open answer)
Authority effect
NOT: to what extent do you agree with the plan on the prime-minister to enforce face
masks in shops?
BUT: to what extent do you agree with the measure on wearing face masks in shops?
Other tips:
Avoid making assumptions
Avoid abbreviations and slang
Avoid negative phrased questions and double negative questions
Use a realistic time frame in your questions
Be aware of researcher effects
Know how to ask questions
The order of the questions
, The type of responses
Standardization of the survey and data collection
Final tips
Check for existing indicators: if established measures exist, use them
Behaviors, attitudes and opinions are complex and best measured with a number
(battery) of questions?
Think about answer categories
o Avoid categorical answers of scales (e.g. age) as much as possible
o Behaviors, attitudes and opinions best answered with a Likert scale
o Avoid yes/no questions as much as possible
o Avoid multi-option responses
Limit the number of questions if possible
Make sure your survey has a coherent ‘story’
Pilot your survey
What needs to be in a survey
Step 1: Introduction
Introduce the topic and the reason this survey is done
Explain the type of questions respondents will get
o But don’t give the exact questions
Informed consent
o Make this mandatory
Contact details of researcher
Ethical issues
o Do not lie or deceive a respondent, but what if ‘innocent deception’ is needed?
See also the video on BS in the module ‘How to make a survey’
Step 2: Your questions
Present your survey questions in a logical order
That are unambiguous, clear, etc.
Make sure all questions require an answer (forced response)
Step 3: Ending
End your survey with a ‘thank you’
Offer respondents the opportunity to see the end result
Give your contract details one more time
Final information on the group paper
Choose a topic with your peers, turn your topic into a RQ and turn your RQ into
operationalizable concepts
Make your own (online group) survey
o Use Qualtrics: http://leidenuniv.eu.qualtrics.com
Have 50 respondents per group member fill in your survey
o Can be a convenience sample
Report your findings in a 1000 word paper
, o No 10% extra word limit
o Deadline: Tuesday October 5th at noon
o Hand in via Brightspace
Writing it down
Introduction:
Introduce the topic
How you went from your topic to a general RQ
Discuss theory (if relevant)
Methodology
If you selected pre-existing measures, describe why and reference them. If you made
your own questions, write down how you operationalized them (don’t forget to
reference)
Describe sampling strategy
Be as detailed as possible
Critically reflect on your survey and data collection strategies
To what extent is the survey/data collection reliable and valid?
Lecture 2, 13-09-2021, How to distribute a reliable and
valid survey
A critical eye
Science vs pseudoscience
Preconceived notions of what to believe
Searching for evidence to support a preconceived belief
Ignoring evidence that would falsify a claim or belief
Disregarding alternative explanations for a phenomenon
Accepting flimsy evidence to support an extraordinary claim
Relying heavily on anecdotal evidence
Employing very little skepticism
But is all science good?
Andrew Wakefield:
- In a study of 12 children Wakefield claimed that the measles-mumps-rubella vaccine
caused autism.
- This study still has its grips on society today
- BUT 12 years later his work has been retracted, because it was a sample size he only
looked at 12 children
o Very small sample size = 12
o Cause Effect, control group; did not have that
o Falsified outcomes
Always remain critical, not all research is quite good.
- Drinking wine before bed could help you lose weight.
