Week 1 Flashcards
Psychometric Theory
The measurement or quantification of psychological constructs through various instruments or psychological tests! It relates to instrument development
Abnormal
Contextual analysis of something out of the social morals
How do we measure psychological phenomenon?
Indirectly: because some phenomena are too prevalent to do directly an example of indirect would be certain scales used.
Do we directly measure psychological phenomenon?
No directly only occurs in the medical field such as getting blood tests or Covid tests.
What are the three reasons why psychologists are dependent on patients?
-dependent on patients to tell us their symptoms, be good historians
-tell us the truth
-patients need to be self-aware of their problems
What happens to the accuracy when we depend on patients?
It goes down.
Why do we want to understand test construction within psychometric theory?
Because you will inherently know the strengths and weaknesses and therefore it can be applied appropriately
Ex: knowing when a test gives a false positive or a false negative
What is measurement?
Psychometric has two primary purposes for measurement 1. Scaling & 2. Classification
Scaling
the process of transforming data values so that they fit within a specific range
Why do we use a scale?
Comparing scales from one person to another, consolidating data! Allows us to standardize, it is objective it’s numerical
Classification
Define whether the objects falls in the same or different categories with respect to a given attribute
Ex: grading scale 93%-100% A
Why do you need to be careful with classification?
It is reductionistic, a person may be classified into a category that does not represent what they actually or truly know, they may be misclassified! Once you classify someone they stay with that diagnosis
How do psychological theories become testable?
Operationally define the psychological construct & develop an instrument to measure the areas that define the construct
What is the hallmark of good measurement?
That there is some form of standardization!
Why do we use standardization?
To get the same results
How is a measure able to be standardized?
Its rules are clear, it is practical to apply, it does not demand great skill of administrators, its results do not depend upon the specific administrator
What is the hallmark of science?
The hallmark of science is replication to get the same results that is why we use standardization.
Measurement of attributes
Attributes are a particular feature of an object! Do not measure objects but the attributes or properties of those objects
What is the problem with with attributes?
We don’t always have common definitions of the attributes.
Advantages of Standardized Measures
-Objectivity, Quantification, Communication & Economy
-these underlie most important aspect of science replication and generalization of scientific findings
Objectivity
Any statement of fact by one scientist should be independently verifiable by other scientists (ex: Freud theory vs Behavioral theory)
Quantification
Numerical details can be reported in greater detail than personal judgments & it permits the use of more powerful methods of understanding like mathematics
Communication
Is facilitated because the language becomes more uniform (through others that work together it is easy to talk to one another) most disciplines do not use the same language but most speak science
Economy
More economical of time and money than are subjective evaluations (it is faster to give number instead of wasting time)
How do scientists develop measures?
By stating the rules to quantify attributes of real objects
Inferential statistics
You take sample data and you make inferences about the population. This sample looks this way based on that population, we don’t use this in psychometric theory
Descriptive statistics
Look at trends in the sample and understand them. Based on the sample itself
What statistics do we use for psychometric theory?
We use descriptive & sampling content
Sampling content
Looking at the items on a test
Measurement scales
Nominal, ordinal, ratio, interval
Nominal
Looking at categories
Ex: what is your favorite type of music?
-Nominal scales can only be coded for frequency data
Ordinal
A set of object is ordered from “most” to “least” with respect to attribute (rank order) units have order, one does not know how much any of the objects possess of the attribute
Ordinal
A set of object is ordered from “most” to “least” with respect to attribute (rank order) units have order, one does not know how much any of the objects possess of the attribute
What is a limitation of Ordinal
Can lose a great deal of information
Interval
Most used in psychology, rank ordering of an object is known, the distances among the attributes is known, the absolute magnitude of the attribute is not known, no absolute 0
Ratio
We do not use these in psychology, because there is no true zero, is an interval scale with a rational or true zero
Why are there very few ratio scales in psychology?
Since individuals do oftentimes possess some of the attribute we are measuring, there is never zero weight ex: weight, height
