Research Methods Exam #1 - Chapter 3

Three types of claims

Frequency, Association, causal

Frequency Claim

Claims about the rate at which something occurs (percentages, fractions). Variables always measured. Only measures one variable.

Association Claims

Claims that one level of a variable is likely to be associated with a particular level of another variable. Has two variables

Causal Claims

Claims that one variable is responsible for changing the other variable. Only have two variable. Uses verbs that suggest one variable "causes" the other

Four Validities

Construct, External, Statistical, Internal

Construct Validity

How well variable is measured and manipulated. How well researcher has operationalized each variable

External Validity

Do results apply to other populations, settings, and contexts? How well is the population represented?

Statistical Validity

Do the numbers support the claim(s)? What is the margin of error?

Internal Validity

Only used with Causal claims

Reliability

Symptom that measure is relatively error free. A good measure should provide the same result upon repeated administrations

Classical Test Theory

NOT a means/way to improve reliability

NOT a way to demonstrate reliability

IS a way to understand what a test score is

Observed Score

Score you get on a measurement (test score, height)

True Score

True underlying standing on the trait (perfect measure)

Error

Anything else that influences observed score. Random

Why is reliability important?

Signals that measurement is RELATIVELY ERROR FREE

3 types of reliability

Test-retest, interrator, internal

Test- Retest Reliability

Give test twice and look for similar results each time/small amount of error

Interrator Reliability

Two raters generate similar ratings, correlation .8 or above

Internal Reliability

Ratings are consistent within individuals. Cronbachs alpha (average correlation) ex. measuring life satisfaction

Characteristics of good measures

Reliability, validity, interrogating

Two categories under construct validity

Reliability and validity

Reliability

Test retest, interrator, internal

Validity

Construct, face, content, criterion

Three criteria for causal claims

Covariance, Temporal precedence, internal validity,

Temporal precedence

Comes first in time before other variable. "music lessons enhance IQ" study must show that music lessons came first and gains in IQ came later

Covariance

first criterion a study must satisfy in order to establish a causal claim. Two variables need to be related

Internal Validity (one of three criteria for causal claims)

Study should be able to eliminate alternative explanations for the association

Independent Variable

Manipulated Variable

Dependent Variable

Measure Variable

Operationalize

Turn a concept of interest into a measured or manipulated variable

What is the difference between a variable and its levels?

Basically, the number of levels of an independent variable is the number of experimental conditions. I.e: if the variable is coffee the levels is the # of cups drank

Explain why some variables can only be measured, and not manipulated

No, because of ethical and practical constraints. I.e: you can't turn someone into a smoker or you can't change someone's age

What is the difference between the conceptual variable and the operational definition of a variable? How might the conceptual variables " affections", "intelligence", and "stress" be operationalized by a researcher?

A conceptual definition tells you what the concept means. Specifically defining a specific concept (variable) so it can be measured. An operational definition only tells you how to measure it.

I.e: High self-esteem might be conceptually defined as a person demonstrating a high degree of self-worth. Operationally, you might define it as scoring above a certain number of a self-esteem scale.

Validity

Refers to the appropriateness of a conclusion or decision, and in general, a valid claim is reasonable, accurate, and justifiable.

How many variables are there in a frequency/Association/causal claim?

Frequency claims describe a particular rate or degree of a single variable. Association claims argue that one level of a variable is likely to be associated with a particular level of another variable and it must involve at least two variables. A causal claim argues that one if the variables is responsible for changing the other. Two variables must be present.

How can the language used in a claim help you differentiate between association and causal claim?

Causal claims use language to suggest that one variable causes the other such as cause, enhance, helps, leads to, adds, increases, and curbs. Association claims use words like at risk for, predicts, is tied to, correlated with, likely, prefers.

How are causal claims special, compared with the other two claims?

They go beyond a simple association between the two variables. They use language to suggest that one variable causes the other. They make a stronger statement

What are the 3 criteria that causal claims must satisfy?

1. Must establish that the 2 variables are correlated

2. Show that the causal variable came first and the outcome variable came later.

3. Establish that no other explanations exist for the relationship

Which of the 4 big validities should you apply to a frequency/association/and causal claim?

Construct Validity

What is internal validity? Why is it mostly relevant for causal claims?

A study should be able to eliminate alternative explanations for the association. Important for causal claims b/c one of the criteria for causal claims is establish no other explanations exist for the relationship to exist.

Define external validity, using the term generalize in your definition

The extent to which the results of a study generalize to some larger population.

Describe at least 3 things that statistical validity addresses

1. How strong is the association

2. The statistical significance of a particular association

3. What is the margin of error of the estimate

What question(s) would you ask to interrogate a study's construct validity?

How well were the variables measured or manipulated?