Quantitative and Qualitative Research Techniques in the Social Sciences
1. Introduction to the course
Techniques for emperical research in social sciences
- Lots of empirical research to understand phenomena
- From the daily newspapers → impact of continuous availability of movies and TV-series on stress,
reporting ill at work and feeling guilty?, do countries gold reserves protect against inflation?, do
dutch households waste less food than before?
- Both theory (to further scientific knowledge) and practice-oriented (to improve the worlds
/companies/ goverments/ ngo’s well-being)
Established, agree-upon ways/tools by which we aim to get reliable (replicable) and valid (saying
something about the thing you want to say something about) results
- Add (scientific) rigor
- So that we ourselves will have confidence in the reliability and validity of our results
- So that others will have confidence in our results
- As other you should be eager (gretig) and able to understand and critically appraise the steps
that were taken
It ain’t only gold that shines out there
De p-waarde of overschrijdingskans (van een gegeven steekproefuitkomst)= de kans dat in de
verdeling gegeven door de nulhypothese de waarde van de toetsingsgrootheid wordt behaald of
overschreden (links, rechts dan wel tweezijdig). De p-waarde is dus gebaseerd op de specifieke
steekproefwaarde. De nulhypothese wordt verworpen, als de p-waarde kleiner is dan het vooraf
vastgestelde significantieniveau, alfa.
Your position in this
You shouldn’t want to be:
- Author of a number of these original scientific articles (you lose your good reputation)
- Author of an applied-research report for of which the findings have low reliability (you spoil their
money, lose your job)
- Someone who just trusts and takes decisions based on the conclusions from emperical studies
with low reliability (lose your good reputationk, money, job)
Black people more red cards?
Different techniques, different controls, different results
It ain’t easy
Annemarie Mol
- Scientists should only present their results but also their method – the outcome of a study is
always linked to the method
- When asked about the value of the results of a controversial study → I haven’t read the article,
so I don’t know. More generally however, also in this case you want to know about het method
Methods and techniques in your study program
- Previous cources BBC students (part of learning trajectory on research methods)
- YRM & MAT 1 en MAT 2
,Learning outcomes
- Explain techniques
- Apply techniques
- Appraise results from techniques
- Necessary knowledge and skills … not only to do your own research … but also critically appraise
research done by others (part of a scientific attitude)
Quiz
What can one say about the typical number of observation units (!) in the three different research
desings?
A. Experiment – case study – survey (cross-sectional)
B. Survey (cross-sectional) – experimental – case study
C. Experiment – survey (cross-sectional) – case study
D. Suvery (cross-sectional – case study – experiment
What can one say about the typical way of data collection in the three different research desings?
A. Experiment & case sutdy standardized. Survey (cross-sectional) unstandardized / adaptive.
B. Survey (cross-sectional) & case study standardized. Experiment unstandardized / adaptive.
C. Experiment standardized. Survey (cross-sectional) & case study undstandardized / adaptive.
D. Survey (cross-sectional) & experiment standardized. Case study unstandardized / adaptive.
Quantitative or qualitative?
- Research question should be starting point
- Amount of prior knowledge
- Size of the population (small population, probably no use of quantitative research)
- Combination of strategies is often very useful (in parallel, one after the other) → mixed methods
(week 5)
About you … and this course
- Research techniqes are quite generic (and their principles even more so). Though researches
from different disciplines can have different preferences.
- Examples in this course are very diverse, but not every example will appeal to everyone and you
may not immediately see the relevance for the kind of things that you are interested in
Quantitative reserach: questions/ideas/models
- What is the effect of trust in institutions and organizations have on consumer confidence?
- What is the effect of eWOM and internal reference price on tourists willingness to pay for an
accommodation?
- Does the impact of exogenous shocks on Icelandic inbound tourism persist, or disappear
relatively fast?
- Which more fundamental belief constructs cna be distinguished leading to support for a marine
conservation area?
➔ Matching data with models
➔ Book by Andy Field, lectures, assignments (with assistance during computer practicals), tutorials
(additional issues and material, and answers to your questions)
,What to expect: formulas
Provide a powerful, concise and (potentially) unambiguous language to talk about statistical models
and to increase your understanding of what is going on, not to impress or to unnecessarily confuse
you
- Andy: it doenst matter whether you use the symbol a or b for something, as long as you define it
correctly and timely
- Me: I sometimes deviate form Andy’s notation, when I think that it increases consistency and
may help you understanding
Assignments
- Steps in learning process: book, lecture, assignment, tutorial
- Answer all questions, but don’t hand in everything
- You still get a grade (10% of final grade)
- Lenient grading
1 point= 7-10, 0,5 points= 5-6, 0 points= 1-4
Grade= 1 + 3Ass3 + 3Ass5 + 3 Ass6
- Model answers become available on BS
- Compare with you own answers and check your own understanding 9these are your most
important deadlines)
- Exam grading will be less lenient (mild)
- Example of exam questions will become available on BS and discussed
IBM SPSS Statistics
Only a tool, not the focus
Same ingredients
- Explaroratoyr factor analysis
- Factorial analysis of variance
- Oneway analysis of variance
- Multiple regression
- Simple regression
- Correlation
- Basic statistics
, Measurement units and standardization
2. Correlation and simple regression analysis
This course: six/five/three/two/one quantitative techniques
1. Introduction to the course
Techniques for emperical research in social sciences
- Lots of empirical research to understand phenomena
- From the daily newspapers → impact of continuous availability of movies and TV-series on stress,
reporting ill at work and feeling guilty?, do countries gold reserves protect against inflation?, do
dutch households waste less food than before?
- Both theory (to further scientific knowledge) and practice-oriented (to improve the worlds
/companies/ goverments/ ngo’s well-being)
Established, agree-upon ways/tools by which we aim to get reliable (replicable) and valid (saying
something about the thing you want to say something about) results
- Add (scientific) rigor
- So that we ourselves will have confidence in the reliability and validity of our results
- So that others will have confidence in our results
- As other you should be eager (gretig) and able to understand and critically appraise the steps
that were taken
It ain’t only gold that shines out there
De p-waarde of overschrijdingskans (van een gegeven steekproefuitkomst)= de kans dat in de
verdeling gegeven door de nulhypothese de waarde van de toetsingsgrootheid wordt behaald of
overschreden (links, rechts dan wel tweezijdig). De p-waarde is dus gebaseerd op de specifieke
steekproefwaarde. De nulhypothese wordt verworpen, als de p-waarde kleiner is dan het vooraf
vastgestelde significantieniveau, alfa.
Your position in this
You shouldn’t want to be:
- Author of a number of these original scientific articles (you lose your good reputation)
- Author of an applied-research report for of which the findings have low reliability (you spoil their
money, lose your job)
- Someone who just trusts and takes decisions based on the conclusions from emperical studies
with low reliability (lose your good reputationk, money, job)
Black people more red cards?
Different techniques, different controls, different results
It ain’t easy
Annemarie Mol
- Scientists should only present their results but also their method – the outcome of a study is
always linked to the method
- When asked about the value of the results of a controversial study → I haven’t read the article,
so I don’t know. More generally however, also in this case you want to know about het method
Methods and techniques in your study program
- Previous cources BBC students (part of learning trajectory on research methods)
- YRM & MAT 1 en MAT 2
,Learning outcomes
- Explain techniques
- Apply techniques
- Appraise results from techniques
- Necessary knowledge and skills … not only to do your own research … but also critically appraise
research done by others (part of a scientific attitude)
Quiz
What can one say about the typical number of observation units (!) in the three different research
desings?
A. Experiment – case study – survey (cross-sectional)
B. Survey (cross-sectional) – experimental – case study
C. Experiment – survey (cross-sectional) – case study
D. Suvery (cross-sectional – case study – experiment
What can one say about the typical way of data collection in the three different research desings?
A. Experiment & case sutdy standardized. Survey (cross-sectional) unstandardized / adaptive.
B. Survey (cross-sectional) & case study standardized. Experiment unstandardized / adaptive.
C. Experiment standardized. Survey (cross-sectional) & case study undstandardized / adaptive.
D. Survey (cross-sectional) & experiment standardized. Case study unstandardized / adaptive.
Quantitative or qualitative?
- Research question should be starting point
- Amount of prior knowledge
- Size of the population (small population, probably no use of quantitative research)
- Combination of strategies is often very useful (in parallel, one after the other) → mixed methods
(week 5)
About you … and this course
- Research techniqes are quite generic (and their principles even more so). Though researches
from different disciplines can have different preferences.
- Examples in this course are very diverse, but not every example will appeal to everyone and you
may not immediately see the relevance for the kind of things that you are interested in
Quantitative reserach: questions/ideas/models
- What is the effect of trust in institutions and organizations have on consumer confidence?
- What is the effect of eWOM and internal reference price on tourists willingness to pay for an
accommodation?
- Does the impact of exogenous shocks on Icelandic inbound tourism persist, or disappear
relatively fast?
- Which more fundamental belief constructs cna be distinguished leading to support for a marine
conservation area?
➔ Matching data with models
➔ Book by Andy Field, lectures, assignments (with assistance during computer practicals), tutorials
(additional issues and material, and answers to your questions)
,What to expect: formulas
Provide a powerful, concise and (potentially) unambiguous language to talk about statistical models
and to increase your understanding of what is going on, not to impress or to unnecessarily confuse
you
- Andy: it doenst matter whether you use the symbol a or b for something, as long as you define it
correctly and timely
- Me: I sometimes deviate form Andy’s notation, when I think that it increases consistency and
may help you understanding
Assignments
- Steps in learning process: book, lecture, assignment, tutorial
- Answer all questions, but don’t hand in everything
- You still get a grade (10% of final grade)
- Lenient grading
1 point= 7-10, 0,5 points= 5-6, 0 points= 1-4
Grade= 1 + 3Ass3 + 3Ass5 + 3 Ass6
- Model answers become available on BS
- Compare with you own answers and check your own understanding 9these are your most
important deadlines)
- Exam grading will be less lenient (mild)
- Example of exam questions will become available on BS and discussed
IBM SPSS Statistics
Only a tool, not the focus
Same ingredients
- Explaroratoyr factor analysis
- Factorial analysis of variance
- Oneway analysis of variance
- Multiple regression
- Simple regression
- Correlation
- Basic statistics
, Measurement units and standardization
2. Correlation and simple regression analysis
This course: six/five/three/two/one quantitative techniques
