Lecture 26: Cirtical Thinking About Psychological Measurement Flashcards
Who was Francis Galton
He was the first person to think about measuring cognition and intelligence and individual differences —> first time people started thinking of psychological traits as quantitative (measurable)
What is positive manifold, who came up with it and what is the mechanism behind it
Charles spearman proposed the idea that people that are better at arithmic are also better at verbal etc tasks (they are all positively correlated), the proposed mechanism behind this is general intelligence that people are born with
Who was Norman Campbell
He was a physicist that believed that measurement is a result of a process of assigning numbers such that each object is represented by a single number and that concatenation operation is possible
what is concatenation operation
The sum of two assigned numbers represents the result of an empirical combination of objects
T/F: numerical relations mirror empirical relations
True, this is a very important fact and most modern technology is based on this
What is the representationalist solution and who cam up with it
Stevens; The fact that measurement is the assignment of numerals according to a rule (any rule), and that the rule followed determines the scale levels, scale levels are defined by the set of transformations that leave the isomorphism intact
What 4 scale levels are there, what do they mean and what are their admissible transformations
- Nominal; number only represent equivalence. Two objects that get the same number have the same property. —> all one-to-one transformations are admissible
- Ordinal; number represent order. Objects that are assigned higher numbers have more of the property represented. —> all monotonic transformations are admissible
- Interval; numbers represent order, but distances between assigned numbers also have a definite meaning that is the same at all levels of the scale. The zero-point is arbitrary. —> all transformations that preserve order and distances between number are admissible (linear transformations X’=a+bX, where b is a positive constant)
- Ratio; numbers represent order, distances between assigned numbers have meaning and the ratio of numbers have meaning as well. Zero-point is not arbitrary but indicates absence of the property. —> all transformations that preserve the ratio between numbers are admissible (X’=bX, where b is a positive constant)
T/F: parametric tests such as t-tests and ANOVA are not sensitive to transformations that change the distance between scale points
False, they are sensitive and as a result of that, nonlinear transformations will change one’s conclusions
What does the concept of robustness mean
It means to transform your data according to transformations you think should matter, then rerun the analyses. If the conclusions don’t change then they are robust, if they do change that you should investigate why and reconsider your scale levels
What are the 3 properties of measurement levels and what do they entail
- Property of identity; sameness vs differentness - sorting people/objects into categories
- Property of order; conveys info about the relative amount of an attribute people possess, numerals indicate the rank order of people relative to each other
- Property of quantity; numerals provide info about the magnitude of differences between peoples
What two meanings does the number 0 have
- Zero reflects a state in which an attribute has no existence —> absolute zero
- Zero is an arbitrary quantity of an attribute —> arbitrary zero
What are 3 requirements for sorting people into categories
- People in the category must satisfy property of identity
- Categories must be mutually exclusive (you are either in or the other, not both)
- Categories must be exhaustive (which means that there is only the amount of categories that you are using, there cannot be a different category within the dimension of whatever you are measuring that is not included)
