Review Concepts
Measurement Concepts
Validity
Validity Definitions
Reliability
SPSS
100
Which is relating to every possible individual of interest: The population or the sample?
The Population
100
When is the best time to use the median as an alternative to the mean?
When your data include outliers
100
Of the three main types of validity discussed, which of them is concerned with ensuring that the new measure is tapping into the latent construct it is supposed to?
Construct Validity
100
The type of validity that asks, "Do these items look right?"
Face Validity
100
What is the best synonym for "reliability" as discussed in class?
Consistency
100
Based on this output, how many participants completed the math anxiety measure?
338
(N = 338)
200
When can inferential statistics prove a claim?
Never
200
In statistical terms, what is the difference between a constant and a variable?
A constant does not vary; a variable varies
200
Of the three main types of validity discussed, which of them is concerned with comparing the new measure to an existing benchmark?
Criterion Validity
200
The type of validity that asks, "Were these statistics done correctly to justify this conclusion?"
Statistical Validity
200
Of the three main types of reliability discussed in class, which one did we discuss as often being the least reliable (e.g., having the lowest alphas)?
Inter-rater Reliability
200
Based on this output, is math anxiety related to math self-concept, and if so, describe the relation.
Yes. Pretest math anxiety (Pretest SIMA) and pretest math self-concept (MSC) are strongly, negatively related. Meaning that higher values of math self-concept are associated with lower levels of pretest math anxiety.
300
What is this symbol and how is it read?
Capital letter Sigma
It is read, "The sum of..."
300
What type of data is appropriate to graph with a histogram?
Continuous
300
Of the three main types of validity discussed, which of them is concerned with ensuring that the new measuring successfully accesses all components of the latent construct?
Content Validity
300
The type of validity that asks, "Can these results be generalized to the population of interest?"
External Validity
300
You are creating a new scale and you find the internal consistency to be .74. How should you interpret this value?
This meets the given cutoff of > .70, but this alpha is still pretty low. I would look back at the items and see what could be revised.
300
Based on this output, what is the Pearson's r value between "rushing" and "SIMA" ?
-.06
(With a corresponding p value of .273 -- a nonsignificant relation)
400
Imagine you are interested in seeing if the number of children in a home affects the happiness of the parents. Is the variable "number of children in the home" continuous or discrete?
Discrete
(You cannot have fractional children)
400
If most of the scores on a continuous variable are clustered near the left (lower) side of a distribution, how would you describe that distribution in terms of skew?
Skewed right (or positive skew)
400
A test score is meaningless until.... _____.
...you draw inferences beyond the score based on the underlying proposed use of the test or instrument.
400
The type of validity that asks, "Does this measure correlate with other conceptually similar measures?"
Convergent Validity
400
According to classical test theory, X = T + e. What do these letters represent?
X = Measured (raw) score
T = True score
e = Error
400
Based on this output, what is the inter-item reliability for the FIGS scale from these data?
alpha = .91
(Strong reliability)
500
What is the difference between interval and ratio scales of measurement?
The ratio scale includes an absolute or meaningful zero point
500
Which measure of spread is standard deviation, squared?
Variance
500
What do we call the score on a criterion corrected for its unreliability?
The True Criterion Score
500
The type of validity that asks, "Are there any confounds in making conclusions about this measure?"
Internal Validity
500
What does it mean to say that reliability is a necessary but not sufficient component of validity?
A measure cannot be valid unless it is reliable (necessary), but knowing that a measure is reliable does not by itself make the measure valid.
500
Based on this output, what is the inter-rater reliability for the FIGS scale from these data?
Unknown
(This cannot be calculated using this method)