Hoorcollege 1/werkgroep 2
A concept = a basic unit of thinking
Why do we need concepts = to communicate, understand and compare.
Concept formation is situated at the very beginning of your research design process. You can have
contested concepts. E.g. democracy is often conceptualized differently. It is important to have
multiple researchers approve of your concept event though it can be contested.
Problems of non- or ill-defined concepts:
- Bad labeling = respectless/vague
- To vague/to broad. It is too general and loses meaning
- The definition is too small so you limit yourself too much
- No clear explanation.
Difference in kind (either/or) vs. difference in degree (more or less). What is and how much.
- A country is either a democracy or not. Once you have the basic of your concept clear, you
can think about the difference in degree. When you have multiple cases, which one is more
democratic and which one is less democratic.
First you have to have positive and negative concepts clear!
- Positive = a specific attribute which is always present
- Negative = a specific attribute which is always absent.
E.g. democracy is a positive concept while authoritarian regimes are a negative concept. Where do
you put South-Africa? You always have to conceptualize the ‘opposite’ of what you want to do
research on.
Continua = one may span a continuum between the positive and negative concepts.
Classification and typologies:
- Principle of ‘per genus et differentiam’.
o Genus = the class of objects to which it belongs
o Differentiam = the particular attributes that make it different from all other objects in
the same class.
- You can compare cases from the same class, but you cannot compare cases from different
classes.
Classification = an uni-dimensional concept (either/or)
- Purpose: systematizing cases, determining core attributes of a concept.
- Something is either part of an uni-dimensional concept or not. It is either part of the positive
concept or the negative concept.
Typology = an two- or multidimensional concept
- Purpose: systematizing cases, explanation.
- When you compare multiple classes with each other you can make relationship or
connections.
Important characteristics of both classification and typologies:
- Mutual exclusiveness: each case belongs to one class or type only
- Exhaustiveness: each case must belong to one class or type.
Types of systematized concepts = have a structure that determines how attributes are linked to each
other
- Necessary conditions concept (AND (*) concept)
o The case must exhibit all attributes in order to be subsumed under the concept. You
need all conditions to be sufficient for the concept.
1
, - Family resemblance concept (OR (+) concept)
o The case must exhibit one attribute in order to be subsumed under the concept. You
need one condition to be sufficient.
Capital letters are present conditions. Small letters are absent conditions. This is a deterministic
relationship.
This model is used in quantitative papers. X is a variable and beta is the degree of effect. The … are
other variables that may be in the equation. * is multiplication. + is an addition, not an OR. This is a
probabilistic relationship.
Ladder of abstraction/ ladder of generality:
What does this concept consist of:
- Extension = case to be defined, described, ordered
- Intension = attributes defining the case covered
- Label = clear name, unambiguous, neutral.
When you have a lot of attributes, you cannot define a lot of cases. The mor cases you explain, the
less attributes you can use.
2
, Satori’s rules of concept formation:
- Analyze cases only as a part of a wider class of cases
o You first have to look at the universe of cases before looking at a specific case.
- Compare only within classes (or cases).
- If cases do not share one class, move up the ladder of abstraction
- If results are meaningless/too general, move down the ladder of abstraction. More attributes
but less cases.
Conceptual stretching = assigning a case to a (root) concept to which the case does not belong. By
doing this you make your concept less valuable.
Elimination of conceptual stretching = finding/specifying the concept to which the case belongs.
What to do when you have new concepts that do not fall into a class? You cannot just add them
because then you get conceptual stretching.
- Moving up the ladder of abstraction. This leads to loss of differentiation. You lose attributes,
but you can have more cases in a class
- Moving down the ladder of abstraction, classical subtypes (less extension). You get less cases,
but more attributes.
- Specifying diminishing subtypes
- ‘precising’ the definition of the root concept by adding defining attributes
- Lowering the standard of the overarching concept.
Other approaches to concept formation (other than systematized concepts)
- Necessary conditions concept (AND) and family resemblance concepts (OR).
INUS condition = neither individually necessary nor individually sufficient for an outcome. Instead it is
one cause withing a combination of causes that are jointly sufficient for an outcome.
- E.g. your house burns down but this can have multiple causes. However, the outcome of all
causes is your house burning down.
3
A concept = a basic unit of thinking
Why do we need concepts = to communicate, understand and compare.
Concept formation is situated at the very beginning of your research design process. You can have
contested concepts. E.g. democracy is often conceptualized differently. It is important to have
multiple researchers approve of your concept event though it can be contested.
Problems of non- or ill-defined concepts:
- Bad labeling = respectless/vague
- To vague/to broad. It is too general and loses meaning
- The definition is too small so you limit yourself too much
- No clear explanation.
Difference in kind (either/or) vs. difference in degree (more or less). What is and how much.
- A country is either a democracy or not. Once you have the basic of your concept clear, you
can think about the difference in degree. When you have multiple cases, which one is more
democratic and which one is less democratic.
First you have to have positive and negative concepts clear!
- Positive = a specific attribute which is always present
- Negative = a specific attribute which is always absent.
E.g. democracy is a positive concept while authoritarian regimes are a negative concept. Where do
you put South-Africa? You always have to conceptualize the ‘opposite’ of what you want to do
research on.
Continua = one may span a continuum between the positive and negative concepts.
Classification and typologies:
- Principle of ‘per genus et differentiam’.
o Genus = the class of objects to which it belongs
o Differentiam = the particular attributes that make it different from all other objects in
the same class.
- You can compare cases from the same class, but you cannot compare cases from different
classes.
Classification = an uni-dimensional concept (either/or)
- Purpose: systematizing cases, determining core attributes of a concept.
- Something is either part of an uni-dimensional concept or not. It is either part of the positive
concept or the negative concept.
Typology = an two- or multidimensional concept
- Purpose: systematizing cases, explanation.
- When you compare multiple classes with each other you can make relationship or
connections.
Important characteristics of both classification and typologies:
- Mutual exclusiveness: each case belongs to one class or type only
- Exhaustiveness: each case must belong to one class or type.
Types of systematized concepts = have a structure that determines how attributes are linked to each
other
- Necessary conditions concept (AND (*) concept)
o The case must exhibit all attributes in order to be subsumed under the concept. You
need all conditions to be sufficient for the concept.
1
, - Family resemblance concept (OR (+) concept)
o The case must exhibit one attribute in order to be subsumed under the concept. You
need one condition to be sufficient.
Capital letters are present conditions. Small letters are absent conditions. This is a deterministic
relationship.
This model is used in quantitative papers. X is a variable and beta is the degree of effect. The … are
other variables that may be in the equation. * is multiplication. + is an addition, not an OR. This is a
probabilistic relationship.
Ladder of abstraction/ ladder of generality:
What does this concept consist of:
- Extension = case to be defined, described, ordered
- Intension = attributes defining the case covered
- Label = clear name, unambiguous, neutral.
When you have a lot of attributes, you cannot define a lot of cases. The mor cases you explain, the
less attributes you can use.
2
, Satori’s rules of concept formation:
- Analyze cases only as a part of a wider class of cases
o You first have to look at the universe of cases before looking at a specific case.
- Compare only within classes (or cases).
- If cases do not share one class, move up the ladder of abstraction
- If results are meaningless/too general, move down the ladder of abstraction. More attributes
but less cases.
Conceptual stretching = assigning a case to a (root) concept to which the case does not belong. By
doing this you make your concept less valuable.
Elimination of conceptual stretching = finding/specifying the concept to which the case belongs.
What to do when you have new concepts that do not fall into a class? You cannot just add them
because then you get conceptual stretching.
- Moving up the ladder of abstraction. This leads to loss of differentiation. You lose attributes,
but you can have more cases in a class
- Moving down the ladder of abstraction, classical subtypes (less extension). You get less cases,
but more attributes.
- Specifying diminishing subtypes
- ‘precising’ the definition of the root concept by adding defining attributes
- Lowering the standard of the overarching concept.
Other approaches to concept formation (other than systematized concepts)
- Necessary conditions concept (AND) and family resemblance concepts (OR).
INUS condition = neither individually necessary nor individually sufficient for an outcome. Instead it is
one cause withing a combination of causes that are jointly sufficient for an outcome.
- E.g. your house burns down but this can have multiple causes. However, the outcome of all
causes is your house burning down.
3
