Midterm 2 - Interviewing and Child Witnesses (Ch. 8 and Bruck et al.)

Question 1

A meta-analysis of validity of job selection interviews at predicting job performance found a mean validity of ___ for unstructured interviews and ___ for structured ones

Answer 1

.31; .62

Question 2

What are 4 reasons structured job interviews have better validity?

Answer 2

• use standardized Qs
• analysis of job domain (Qs correspond to attribute being assessed)
• well-defined rating scales
• mechanical combination of ratings (assign points to diff responses to remove subjectivity from scoring)

Question 3

What is the SCID?

Answer 3

• structured clinical interview for DSM-V
• standardized Qs corresponding to domain of psychopathology
• decision tree to arrive at diagnosis

Question 4

(T/F) when assessing a client for a DSM diagnosis, the first step is usually to complete the SCID

Answer 4

FALSE, start w unstructured interview to build rapport

Question 5

What are 2 methods used to assess reliability of SCID? Which method would we expect to generate lower reliability coefficients?

Answer 5

• joint interviews (interviewer and observer)
• test/retest design (2 time points, maybe diff interviewers)
• test/retest should have lower reliability bc more sources of error

Question 6

In assessing reliability, a Kappa value of ___ indicates good agreement between raters, ___ indicates fair agreement, and ____ indicates poor agreement

Answer 6

> .70 is good, .50-.70 is fair, <.50 is poor

Question 7

SCID reliability (joint interview and test-retest) is lowest for which 3 disorders?

Answer 7

• Persistent depressive disorder (.70 joint, .46 retest)
• Alcohol use disorder and Panic disorder both have lower retest reliability (.55 and .52) but good joint reliability

Question 8

SCID reliability (joint interview and test-retest) is highest for which 2 disorders?

Answer 8

• Bipolar 1 (1.0 joint, .96 retest)
• Schizophrenia (.97 joint, .92 retest)

Question 9

What is the desirable standard for evaluating SCID validity?

Answer 9

• LEAD (best estimate diagnosis)
• Longitudinal assessment
• done by Expert diagnosticians
• using All Data available ab subject

Question 10

Basco et al assessed convergent validity for SCID and diagnosis following intake interview and found….

Answer 10

• generally low validity
• okay but not great validity (.30ish) for Mood Disorders, Substance Abuse, and Eating Disorders (these might just be easier to pick up on generally)

Question 11

What evidence do we have for content, criterion, and construct validity of the SCID?

Answer 11

• content: close corresp. btw SCID and DSM
• criterion: LEAD may be possible but not well studied
• construct validity: problems w discriminant validity (more of an issue w diagnoses/classification and not SCID)

Question 12

SCID has _____ reliability for many disorders and excellent ___ validity but limited ____

Answer 12

fair to good reliability; excellent content validity; limited other forms of validity

Question 13

What are the 4 major characteristics of interviewer bias?

Answer 13

• a priori beliefs ab occurrence
• collection of confirmatory evidence only
• failure to test alternative hypotheses
• ignore discrepant evidence

Question 14

What are 7 main components of suggestive interviewing?

Answer 14

• info introduced by interviewer that was not mentioned by interviewee
• few open-ended Qs
• leading/misleading Qs
• emotional tone (selective reinforcement, bribes, threats)
• aggrandizement of interviewer status
• visualization procedures/pretending

Question 15

What are the 3 types of questions in the NICHD protocol?

Answer 15

• main invitation (tell me everything that happened from start to end)
• follow up invitation (tell me more ab that)
• follow up and cued invitations (earlier you mentioned X, tell me everything about that)

Question 16

What are the 10 stages of the NICHD protocol?

Answer 16

1. Introduction
2. Establish ground rules for truthfulness
3. Build rapport w interviewee
4. Conduct a practice interview for memory training
5. Transition to substantive phase
6. Open-ended prompts
7. Ask focused Qs ab undisclosed info followed by open-ended prompts
8. Discuss disclosure information
9. Conclude; invite additional info or questions
10. End w neutral topic for closure

Question 17

(T/F) the first step of the NICHD protocol is asking “Wh” questions

Answer 17

FALSE, only ask them after exhaustive free-recall prompting

Question 18

What are 3 things the NICHD protocol suggests to do only after exhaustive free-recall prompting?

Answer 18

• directive Qs (“wh”)
• option-posing Qs (yes/no Qs)
• suggestive utterances strongly discouraged at any time!

Question 19

What are the main findings of Bruck et al.’s study?

Answer 19

• preschoolers, 2 true and 2 false events (pos and neg) presented over 5 suggestive interviews
• after only 2 suggestive interviews, assent rates were similar for true and false events
• false narratives contained more details, more markers of narrative cohesiveness, and more elaborations
• decreases in accuracy accompanied by increases in markers of credibility!!