PAID (BESC1534): Reliability & Validity

Validity is about [...] and Reliability is about [...]

• Accuracy; Authenticity

• Consistency; Accuracy

• Consistency; Stability

• Accuracy; Consistency

The degree of inter-relationship among items of a scale (i.e., whether they measure the same construct) is reflected in:

• Person's r

• Cronbach's α

• Cohen's κ

• None of above

A scale achieved a Cronbach's alpha of 1, with item-total coefficients ranging from .20 to .60. Those results are positive and there is no need for further reliability check.

• True

• Depends

• False

Two scales with different lengths achieved the same level of Internal Reliability (e.g., α = .80). All other factors held constant, which test was more effectively developed?

• The shorter scale

• The longer scale

• None of them

• Cannot say

A scale is administered twice on the same population at a two-week interval. The Person's r must be [...] for the good Test-Retest Reliability.

• .60 - .69

• .70 - .79

• .80 - .89

• ≥ .90

For Inter-Rater Reliability, more consistent results among the raters lead to Cohen's kappa coefficient:

• Being positive

• Being closer to 1

• Being closer to 0

• Being negative

A pool of items was developed, and experts were asked to rate the items' necessity. Given:

the ratings for item 5 as in the image;
formula is CRV = (ne – N/2) / (N/2);
ne = number of experts indicating “essential”;
N: Number of experts;

The Content Validity Ratio of item 5 is:

• .20

• .40

• .60

• .75

A school psychologist wants to know if a school entrance exam can predict future academic performance. The students' entrance exam result is compared with their first-semester GPA (because GPA is an accepted criterion variable of academic performance) to see if the entrance exam is valid.

This is an example for:

• Criterion
  Concurrent Validity

• Criterion
  Predictive Validity

• Construct
  Convergent Validity

• Construct Divergence Validity

CFA (Confirmatory Factor Analysis) is used to measure:

• Content Validity

• Criterion
  Concurrent Validity

• Criterion
  Predictive Validity

• Construct Validity

Autism and social anxiety have some similar manifestations but are theoretically different. Therefore, you expect a weak link between your newly developed autism scale and the Liebowitz Social Anxiety Scale. You distribute both questionnaires to a large sample, and based on a very weak correlation between the results, you confirm that your scale has discriminant validity.

This is an example of:

• Content Validity

• Criterion Validity

• Construct
  Convergent Validity

• Construct
  Divergent Validity