Home > Codebook > Measures > Psychopathy Checklist Youth Version (PCL-YV) - Subject Baseline

Psychopathy Checklist Youth Version (PCL-YV) - Subject Baseline

This measure appears in the following time-points: Baseline.

Related Construct

Description of Measure

The Psychopathy Checklist: Youth Version (PCL-YV; Forth, Kosson, & Hare, 2003) assesses psychopathic characteristics among youth. Procedures advocated by the authors suggest a 60-90 minute, semi-structured interview to assess the youth's interpersonal style, obtain information on a variety of aspects of his/her history and current functioning, and assess the credibility of his/her statements. However, the Pathways study was unable to accommodate an interview of this length to assess psychopathy. Instead, all questions from the PCL-YV interview guide were incorporated into the Pathways baseline interview battery. Nearly all were asked in the open-ended format suggested by the PCL-YV authors. Following the baseline interview, interviewing staff generated a report which pulled all of the PCL-related answers from the full interview. With this information in hand, as well as information from court records and the parent collateral interview, the interviewer completed the PCL-YV rating form. This form lists 20 separate items for which the youth is rated on a 3-point ordinal scale: "0" item does not apply to the youth, "1" item applies to a certain extent and "2" item applies to the youth.

The authors of the adult versions of the Psychopathy Checklist have reported a 2-factor structure underlying the item scores (see Forth et al, 2003, page 57); however, the PCL-YV has not replicated this factor structure. Instead, factor analytic work with the PCL-YV has produced both a 3-factor and a 4-factor structure that are acceptable to the PCL-YV authors, each with their own caveats.

Using data from the Pathways sample, Jones et al. (2006) found that correlating items from the PCL-YV increase model fit. Specifically, when estimating the Cooke and Michie (2001) three-factor model, the impression management (Item 1 on the PCL-YV) and grandiose sense of self worth (Item 2 on the PCL-YV) should be allowed to correlate. When testing the four-factor model proposed by Hare (2003), the impression management and grandiose sense of self worth items should be allowed to correlate as well as the serious criminal behavior and criminal versatility items (Items 18 and 20 on the PCL-YV respectively). With these modifications to the three- and four-factor models, we conducted a confirmatory factor analyses with both the three- and four-factor solutions. The three-factor model (Cooke & Michie, 2001) had excellent fit to the data: CFI: .94, TLI .92; RMSEA .06. The four-factor model (Hare, 2003) also showed good fit to the data: CFI: .92, TLI: .89; RMSEA: .06. It is recommended that anyone wishing to work with the factor scores of the PCL-YV read this article before proceeding.

This CFA was only conducted on the male sample (N=1,170).

The inter-rater reliabilities for the separate ratings are not all acceptable (see details below). As a general rule, then, these individual ratings should not be used as separate variables. Only the factor scores and total scores appear to be reliable. These scores were found to have good internal consistency (alpha: Total Score = .87; Factor 1-Interpersonal/Affective = .76; Factor 2-Socially Deviant Lifestyle = .78). Intraclass Correlation Coefficients (ICC) for the factor scores and the total score were acceptable as well: ICC Total Score = .92; ICC Factor 1= .79; ICC Factor 2 = .93

Because of the complexity of administering the PCL:YV, all interviewers completed extensive training. To assess inter-rater reliability during training, we computed intraclass correlation coefficients (ICCs) with a two-way mixed effects model, with raters as a fixed factor and agreement defined as absolute using raters' scoring of four videotaped cases. Our analyses indicated excellent rates of agreement for total scores (ICC .91). As past research has suggested (Jones et al., 2006), we correlated residual errors for Items 1 and 2 (impression management and grandiose sense of self-worth, respectively) and Items 18 and 20 (serious criminal behavior and criminal versatility, respectively) and allowed Item 13 (lacks goals) to load on both the behavioral and affective factors. Our three-factor model (Cooke & Michie, 2001) had the best fit to the data (comparative fit index [CFI] .93, Tucker-- Lewis index [TLI] .92; root-mean-square error of approximation [RMSEA] .06), but the four-factor model (Hare, 2003) also exhibited good fit (CFI .91, TLI .89; RMSEA .06). In addition, the four factors were significantly correlated (rs ranging from .39 to .59, with all ps .001).

There are six computed scores for this scale:

Note: 3 items (pcl11, pcl17 and pcl20) are not classified and therefore do not appear in the factor scores but are counted in the total score.

The following individual items are also available:

Prorated scores

Sometimes there is insufficient information to score a particular item. In these situations, an item can be omitted. The PCL-YV can accommodate up to five missing items without invalidating the total score. Two items can be missing without invalidating the factor scores. If items are omitted, score are prorated using tables provided by the authors.

Inter-rater reliability

To obtain inter-rater reliability scores, all Pathways interviewing staff watched six taped PCL-YV interviews. Following review of the tape, each staff member completed the PCL-YV rating form for the individual on tape. These scores were then compared to the scores assigned to these same individuals by one of the PCL-YV authors, Adelle Forth. As seen from the chart below, the consistency of scoring in this exercise was not high enough across all the scales to warrant confidence in the use of these ratings.

Domains with Weighted Kappas

Data Issues

There are 54 caseID's that are missing the PCL. We are unable to explain why the PCL is missing for these ID's. These cases have been given a missing value code of '-99 - PCL Missing: Unknown Reasons'.

References