CHAPTER 5 Flashcards
What are two important features of concepts?
The are non-empirical
- don’t have qualities like tate, sound, smell, touch
They’re meaningful
- meaning is what gives concepts life
What is conceptualization? - what is a problem with it?
What is operationalization?
conceptualization = The refinement and specification of abstract concepts (ex. 5 different pens in front of you but you name them all “pens”)
–> problem with conceptualization is that you don’t know if a concept has the same meaning to you and others
operationalization = The development of specific research procedures (operations) that will result in empirical observations representing those concepts
What is a conceptual definition?
What is the other name for it
What’s an example
Indicates the meaning of an abstract term by expressing it in other abstract terms
- also called nominal definitions or non-numerical attributes / categories
- operate entirely on the abstract level
- useful conceptual definitions informs us with both what is included and excluded from the concept
ex. dictionaries - concept whose meaning you don’t know is expressed in terms of concepts you hopefully do know
According to the Governance of marriage: Parliament has the right to _______ marriage
Provinces have the right to _______ marriage
Define
Solemnize
*at one point in time, provinces could recognize a marriage but the federal government wouldn’t
What was Bill C-28
Civil Marriage Act
changed the definition of marriage
Was “the lawful union of one man and one woman to the exclusion of all others”
Now “the lawful union of two persons to the exclusion of all others”
What are the 5 options for marriage status on the Canadian census?
- Single (never marred)
- Married (including common law)
- Separated
- Divorced
- Widowed
What is an operational definition?
Operational definitions specify a set of steps (operations) for connecting an abstract concept with concrete observations
- how do we count or see or empirically observe something “show me type”
ex. recipe for baking - provides steps for producing (experiencing) the baked good
Important because social researchers are interested in linkages between the abstract and concrete levels of experience
Review - What are variables?
What are we doing with variables now?
They are logical groupings of attributes
Two basic types discussed so far: Numeric (quantitative) and categorical (qualitative)
Now we’re going to break them down into more specific definitions (four levels of measurement)
What are the four levels of measurement?
Nominal, Ordinal, Interval, and Ratio
Describe nominal measures of variable grouping
*categorical variable - merely offers names or labels for characteristics
Possess “Equivalence” –>
Exhaustive : every possible observation can be classified in terms of that variable (ex. might add “other” “prefer not to say” “not applicable” or “don’t know”)
Mutually exclusive : every observation can be classified in terms of a single attribute
**all we can say about two people in terms of a nominal variable is that they are either the same or different
Describe ordinal measures of variable grouping
Equivalence AND Rank Order (attributes can be logically rank-ordered)
*can say that two people are the same or different AND can say one is “more” than the other
*we don’t know anything about the distance between each rank though (irrelevant how close or far apart the groups are from one anohter)
ex. social class, political orientation, self-rated health, course grade
Describe interval measures of variable grouping
Equivalence AND Rank Order AND Equidistance (fixed and equal space between groups)
- relatively uncommon in social sciences
*the logical distance between attributes can be expressed in meaningful standard intervals
**we can say two ppl are different from each other, one is more than the other, and HOW MUCH more
- ex. temperature - degrees celsius (can have negative degrees which is important)
Describe ratio measures of variable grouping
Equivalence AND rank order AND equidistance (fixed/equal intervals between attributes) AND a Real Zero (a meaningful or true zero point)
ex. number of hours doing something
ex. age
**can say that two ppl are different, one is more than the other, how much they differ, and the RATIO of one to another
*in most cases, we can treat interval and ratio measures the same
What is the implication of levels of measurement?
The level of measurement is significant in terms of the arithmetic operations that can be applied to a variable
What is Precision vs. Accuracy?
Precision = the property that refers to the fineness of measurement distinctions
Accuracy = the property that refers to the correctness of measurements
ex. guessing teachers’ age - if you’re guessing specific numbers, it’s more precise, but if you’re guessing in grouped ages (50-59), it’s more likely to be accurate
Precision isn’t always desirable but if you can get it then it’s golden and you can do so much with it statistically
What is reliability?
The quality of measurement that suggests the same data would have been collected each time in repeated observations of the same phenomenon
- consistency of measurement
What are the 4 methods for addressing measurement reliability?
- Test-Retest Method
- make the same measurement more than once - Split-Half Method
- give one measurement of the variable to half of your subjects, and the other measurement to the other half
(measuring the same thing a little bit differently and see if you get similar results)
**it’s always good to make more than one measurement of any subtle or complex social concept like prejudice for example - Use established measures
- measures that have proven their reliability in previous research - Reliability of Research Workers
- make use of supervisors or colleagues
What is validity?
A measure that accurately reflects the concept it is intended to measure
*Are you measuring what you want to measure
ex. IQ tests are sometimes considered a valid measure of “intelligence”
What is face validity?
Does it seem to be a reasonable indicator of that concept
- what does your gut tell you just from the face of it
Ex. looking at frequency of church attendance - does this tap into religiosity (yes it seems to make sense without a lot of explanation)
What is criterion-related validity?
Does the measure predict future outcomes?
*AKA, predictive validity
- based on some external criterion
ex. MCAT or LSAT - considered to have good predictive validity - can measure whether you’re going to perform well in medical or law school
What is construct validity?
Does your measure relate logically to other concepts?
ex. If you do have a measure of depression and anxiety - we should see some kind of connection or pattern between the two - if not then maybe your measure of depression is not the most valid
What is content validity?
Does your measure cover the full range of meanings of that concept?
Ex. a unit test doesn’t include anything taught in a whole lecture - not the most valid test of your learning
What is convergent validity?
what’s an example
Using mixed methods (triangulation enhances validity - compensates for the weakness of one method over the other)
ex. hand washing survey
METHOD 1: asked just over 1000 americans on the phone - how often do you wash your hands when using a public restroom → 95% said they always do
METHOD 2: Had research assistants wait in restrooms and observed people - looked at almost 8000 people
Found that 67% of people actually washed their hands
(83% in chicago, 48% in new york → varied by city)
(Varied by gender - 58% of men - 75% of women)
Reliability is a ________ condition for validity
necessary
The validity and reliability of a measure must be weighed against…
- the potential harm
- the potential inconvenience
- the potential intrusion
What is an indicator?
What are composite measures?
An empirical specification of some abstract concept
- using multiple indicators allows for a more adequate capture of the richness of meaningful concepts
**indicators are intended to capture the meaning of the concept - concepts rich with meaning will likely require multiple indicators (ex. indicators for being in love)
- allows exploration into how multiple indicators might be combined into overall, summary measures
Composite measures - adding together or combining different indicators (indexes and scales)
ex. using GPA to measure uni performance
*most common in quantitative methods
What is an index?
A type of composite measure that combines multiple items that, when aggregated, are intended to represent some more general dimension
ex. social class position generator questions:
Do you know a doctor? –> 1 = yes, 2 = no
Do you know a lawyer? –> 1 = yes, 2 = no
Possible score = 2
Doesn’t get at the question of if a doctor is more important or higher social class than a lawyer?
**just accumulating the scores - adding scores together
What is a scale?
what is an example
A type of composite measure composed of several items that have a logical or empirical structure among them
Likert scale : rating scale that measures ppls opinions, attitudes, motivations, etc. - range of answers from one extreme to another
ex. abortion survey
In which cases should abortion be legal?
1. if the mother’s life is at risk
2. If there is a severe anomaly in the foetus
3. If the mother does not want a child
*If somebody answers 3, they know they’re technically answering 1 and 2 as well → the answers’ intensities are forming a pattern - won’t know if they actually form that scale until you analyze the data
Explain how indexes and scales differ?
An index is constructed by simply accumulating scores assigned to individual indicators (they don’t typically need to be in a specific order, and each item is given the same weight in scoring)
A scale is constructed by assigning scores to patterns of responses - recognize that some items reflect a strong degree of the variable while others reflect a relatively weak degree of the variable
What do indexes and scales have in common?
Both are ordinal measures of variables (both rank-order the units of analysis)
Both are composite measures of variables
- measurements based on more than one data item
- indexes and scales are efficient data-reduction devices
what is reification?
the mistake of treating a conceptual construction as something real
what is specification?
using conceptual and operational definitions to clarify the meaning of concepts
Why are “real definitions” useless?
Real definitions are the type used by Eve in the naming of animals (this is a pig because it’s a pig) - they are useless because they assume that there is something intrinsic in a thing that leads to its name
What are dimensions
Specifiable aspects of a concept
*important to be clear about which ones are important in your inquiry
Anomie?
First used before Durkheim to mean disregard for divine law
Durkheim created the social scientific concept of anomie to mean feelings of social dislocation
Describe the “hermeneutic circle” that social researchers sometimes speak of
It is a cyclical process of ever-deeper understanding
The closer determination of the meaning of the separate parts may eventually change the originally anticipated meaning of the totality, which again influences the meaning of the separate parts, and so on.
What is Tautology?
The thinking error that claims to explain something by referring to itself
aka - circular reasoning
*providing explanations requires systematically linking DIFFERENT concepts together
If you were operationalizing attitudes toward nuclear energy, how would you measure the full range of variation?
You’d want to include a range from favouring it very much, through no feelings one way or the other, to opposing it very much
How can you define pragmatism?
“any difference that make no difference is no difference”
Why would you rather to into too much detail compared to too little detail when deciding on the degree of precision in operationalization?
Because when you’re analyzing the data, you can always combine precise attributes into more general categories, but you can never separate out any variations you lumped together during observation and measurement
Every variable must have two important qualities…
exhaustive and mutually exclusive
When your research purposes are not clear, you should seek the __________ level of measurement possible
highest
What was the bielby and bielby study
Looking at variables that predict a screenwriter’s success
Being older (age, ratio), being female (gender, nominal), belonging to an ethnic minority (ethnicity, nominal), and having more years of experience (ratio) were disadvantageous to a screenwriter
Found that reputation (ordinal), rand higher earnings from previous years (measured in ordinal categories) led to more success in the future
*success breeds success
What’s a reliable measure for when you’re asking people for information
Be careful to ask only about things the respondents are likely to know the answer to - things that are relevant to them and be clear in what you’re asking
The _____ variation and richness we allow for a concept, the _______ reliable it is
more
less
Quantitative, nomothetic, structured techniques like surveys and experiments (more reliable or valid??)
Qualitative, idiographic methods like field research and historical studies
(more reliable or valid??)
- reliable
- valid
How are indexes and scales efficient devices for data analysis?
They are efficient data-reduction devices - they allow us to summarize several indicators in a single numerical score, even while very nearly maintaining the specific details of all the individual indicators
Do researchers use indexes or scales more?
Indexes
