Week 2 Flashcards
Defining Concepts
Inductive, deductive and intuition
Inductive reasoning is a way to arrive at a conceptual definition, challenge an existing concept or change an existing concept by using empirical observations first and moving up to forming the concept. In other words, starting and using empirical observations as the basis of the concept
E.g. political regimes; look at different political regimes that have existed and currently exist and would find several characteristics to use this and base definition of political regimes. Bottom up, where there are observations and specific instances used to thus make generalizations about the topic at hand.
Deductive reasoning: reasoning includes having one or more postulates/assumptions and from those having outcomes and effects that logically follow. E.g., if we assume the politicians are utility maximizing in that they try to win the elections in an way - it follows that they would spend campaign funds in the most efficient manner possible - it can also explain the phenomenon where more politicians are moving towards the centre of the political spectrum so that they can appeal to the most amount of people as possible
Intuition: a hunch you might have of a relationship being, for example, you might assume intuitively that conflict leads to less economic growth
Especially if other researchers have not researched the concept yet
Operationalization:
● Is a process in which we move from the abstract to the concrete
● The first level consists of the concept: this is an idea or way of describing observable things and things that exist - such as democracy, it is an idea and way of describing what we observe in some countries and it exists
Variable: a particular way in which variation in the concept is measured:
● The variable is how you translate that abstract concept such as democracy into a quantifiable measure - such as suffrage in this case
● The indicator is the last level and is how we provide a value to our variables - this can be the % of people who are allowed to vote
The instrument used to assign specific values to cases for a given variable:
GPA (variable)
Some use 0-4, 0-10, 0-12
Conversion into numerical grade into alphabetic grade back into a GPA number
This shows us that researchers have different choices on how to measure these variables
Operationalization in practice
Authors looking at the correlation
Concept: political participation - variable: political campaign contributions
Proposition: as political participation increases, social equality increases
Concept: social equality - variable: income equality
Hypothesis: as political campaign contributions increase, income equality increases
Indicators: % of the total population who gave money to a registered political party
Indicators: GINI coefficient - measures income equality from 0-1
Hypothesis: as the percentage of the total population who gives money to political parties increases, the GINI coefficient for that population decreases
There is a narrowing down going on and we have to understand the criteria of this operationalization. Has the researchers also been transparent on these decisions?
Reliability and validity
● Reliability is the consistency of a measure and the likelihood that the results can be repeated using the same methods with a different researcher. A scale is said to be reliable if it measures every person in LB but once we measure some people in KILOS then there is no consistency
● Validity is essentially the degree of fit between your three levels in operationalization - how well do the variables chosen help cover or measure the concept and how well do the indicators chosen help cover the variable. The main type of validity is face/construct validity which asks how logical the decisions at different levels of operationalization are and if the researcher is measuring what they intend to measure
The tradeoff between reliability and validity
There are sometimes tradeoffs between validity and reliability: as we decrease abstraction, we improve reliability, but in doing so, we may decrease validity
Dependent and independent variables
Dependent variable: of particular interest, the outcome you’re trying to explain
Independent variables: the potential cause of variation in an outcome
Ratios
Ratios: (f^1: f^2): can be done by dividing the two numbers with a common number (16,136,930/100,000 = 161 —- 15,475,970/100,000=155 —— 161:155)
f^1 is the number of observations in the fist category
f^2 is the number of observations in the second category
Example 72:62 ratio of men to women in class (1.16)
Rates
Usually a round up number, for example, per 1000 people. For example, there are 511 women per 1000 Canadians. A ratio on the other hand would say there are these many women per these many men. Infant mortality (deaths for every 1000 births), GDP per capita
Percentiles
Percentiles:
The 50th percentile is the median
Quartiles are the number of percentile categories (15th percentile can be 1st quartile, 30th percentile can be 2nd quartile)
Standardization
A variable that has the same unit of measurement as other variables and is therefore appropriate for comparison. Converting all currencies to the US dollar, for example, allows you to compare currencies with a common metric. Standard deviation, rates, and ratios are all examples of standardization.
