Study sets, textbooks, questions
Upgrade to remove ads
MC 3510 Midterm Scholl
Terms in this set (37)
A variable that relates to the IV and DV, creating the appearance of a causal relationship.
: A relationship in which two variables appear to be causally related, but have no causal relationship. The relationship is coincidental or created by a Z variable.
TRUE/FALSE :Just because two variables are related, does not mean one variable causes the other variable
TRUE /FALSE CORRELATION DOES NOT IMPLY CAUSATION
What are the 3 components to To demonstrate a causal relationship between an independent (X) variable and a dependent (Y) variable?
Temporal precedence: X before Y
Covariation: X related to Y
Non-Spuriousness: No other variables involved
A variable whose influence a researcher wishes to eliminate.
Control variable reduces or eliminate the effect of an independent variable on a dependent variable
Graphically: flat lines across X axis; null X->Y relationship
Effect of IV on DV is similar at each value of the control variable
Graphically: lines trend together; positive or negative relationship
Effect of IV on DV varies at different values of the control variable
Graphically: lines cross or one line is flat while other is steep
Relationship between independent and dependent variable depends on the interval or range of the IV being examined.
Increases on the independent variable are associated with a consistent, uniform increase or decrease on the dependent variable.
Concave or Convex, U- or V-Shaped Relationships
Types of curvilinear relationship in which values of the dependent variable increase (concave) or decrease (convex) only in the middle range of the independent variable.
TRUE/FALSE : Median resistant to skew
TRUE FALSE: mean is sensitive to skew, or the degree to which a distribution of a variable leans to one side of the mean
TRUE/FALSE Often better to report the median than the mean, esp. when dealing with highly skewed variables
If a curve leans left, is it skewed positively or negatively? (curve means bell think about the graph!)
If a curve leans right, is it skewed positively or negatively? (curve means bell think about the graph!)
An observation (response) that is numerically distant from the rest of the observations. You can think of outliers as "extreme" observations.
Negatively skewed distributions
have a majority of responses falling on the high end of the scale, with a minority of responses falling on the low end
In negatively skewed distributions, the mean is often dragged in the direction of the skew, to the left
Positively skewed distributions
have a majority of responses falling on the low end of the scale, with a minority of responses falling on the high end
In positively skewed distributions, the mean is often dragged in the direction of the skew, to the right
Mean of interval is simply the average frequency of the response distribution.
Random Measurement Error
Unpredictable fluctuations in variable measurement
- An additive combination of ordinal variables, each of which is coded identically, and all of which are measures of the same underlying concept
Considered a type of "categorical variable."
Responses still take on finite values.
Distance between categories not precisely quantified.
BUT, logical, numerical ordering of categories.
Considered a type of "continuous variable."
Most precise level of measurement.
Distance between response outcomes quantifiable.
Exact, equal distances between outcomes
Considered a type of "categorical variable."
Responses take on finite, categorical values.
No logical ordering of response categories.
Distance between response categories not quantifiable.
When do we need to transform variables?
To creative additive or averaged indices
To collapse a variable, combining its values or codes into a smaller number of useful categories
To recode variables that are coded in an illogical order
Seeks to understand larger truth of individual circumstance
Starts with data, builds theory
Seeks to produce generalizable results
Starts with theory, makes hypotheses, tests them with data
each person is asked the same question in the same way so that any differences between answers are held to be real ones and not the result of the interview situation itself
The content and the context of the interview are both important aspects of the process
Interviewees have the freedom to tell their stories in their own way, although the interviewer may prompt in order to keep the narrative going.
TRUE/FALSE Should qualitative interviewing take place in a natural setting?
Independent Variable(political experiments)
Changes due to things outside your study
Dependent variable = (supposed) effect
Levels (treatments) of the IV
a tabular summary of the relationship between two categorical (nominal or ordinal) variables.
Sets with similar terms
Research Methods Test 3
Research Methods Concept Quiz
SLWK380 - Ch. 6