Conners CBRS Update

The following updates have been made to the Conners Comprehensive Behavior Rating Scales™ (Conners CBRS™):
(1) Validity scale interpretation, (2) T-score interpretation, (3) renaming the Aggressive Behaviors scale to Defiant/
Aggressive Behaviors, (4) re-scoring the Violence Potential scale, and (5) adjusting triggers for the Other Clinical
Indicators. These changes should improve the utility of the assessment in actual practice.

1. Validity Scale It is recommended that clinical judgment be used in the

interpretation of elevated Validity scale scores (i.e., raw
Interpretation scores that are above the cut-offs presented in Table 1).
The Conners CBRS includes three Validity scales: the Item-level analysis can be useful in this process. Elevated
Positive Impression (PI), Negative Impression (NI), and PI and NI scores may indicate a positive response style for
Inconsistency Index (IncX) scales. In the initial release the PI and a negative response style for the NI. However, it
of the Conners CBRS documentation, the guidelines for is possible that these scales could be elevated due to other
interpretation of these Validity scales used the following reasons, such as in the case of an extremely well-behaved
language: probably valid, possibly invalid, and probably youth (an elevated PI score), or in the case of a youth
invalid. To better align the interpretation guidelines who really does misbehave most of the time (an elevated
with the intent of the scales, the guidelines have been NI score). Similarly, while an elevated IncX score likely
revised using new language. Validity scale scores should reflects an inconsistent response style, it is possible that it
be considered as indicative of potentially problematic is the youth’s behavior (rather than the rater’s responses)
response styles (i.e., possible positive, negative, or incon- that is inconsistent. Reviewing the responses to the items
sistent response style; see Table 1). that make up these scales can help determine if there is a
problematic response style or if the scale elevations are
accurate reflections of the youth’s behavior.

Response Style Analysis using the Conners CBRS Validity Scales

Table 1.a Positive Impression
Parent Teacher Self-Report Interpretive Guideline

0–4 0–4 0–3 Overly positive response style not indicated.

Possible positive response style.

5–6 5–6 4–6
Scores may present a more favorable impression than is warranted.

Table 1.b Negative Impression

Parent Teacher Self-Report Interpretive Guideline

0–4 0–4 0–4 Overly negative response style not indicated.

Possible negative response style.

5–6 5–6 5–6
Scores may present a less favorable impression than is warranted.

Table 1.c Inconsistency Index

Parent Teacher Self-Report Interpretive Guideline

0–5 or < 2 absolute 0–4 or < 2 absolute 0–8 or < 2 absolute

differences equal to differences equal to differences equal to Inconsistent responding style not indicated.
2 or 3 2 or 3 2 or 3

≥ 6 and at least two ≥ 5 and at least two ≥ 9 and at least two Responses to similar items showed high levels of inconsistency.
absolute differences absolute differences absolute differences Scores may not accurately reflect the individual due to a careless or
equal to 2 or 3 equal to 2 or 3 equal to 2 or 3 unusual response to some items.

1
 Conners Comprehensive Behavior Rating Scales™ Supplement

2. T-Score Interpretation Note that these guidelines are approximations and should
not be used as absolute rules. There is no reason to believe
A T-score is a standardized score, which means that it can that there is a perceptible difference, for instance, between
be easily compared across different raters or administra- a T-score of 64 and a T-score of 65. Even if a youth
tion dates. T-scores for Conners CBRS scales convert receives a score in a given range, if other information
the raw scores to reflect what is typical or atypical for (e.g., observation, interview, clinical history) suggests
that age and gender. All T-scores have a mean of 50 and something different, then this must be taken into consid-
a standard deviation of 10. A perfectly average T-score eration in the interpretation process.2
of 50 is exactly equal to the mean score of that age and
gender group. The average range falls within one standard This change applies to the interpretation of T-scores
deviation of the mean (i.e., between 40 and 59). For some for the Conners CBRS Content scales and DSM-IV-TR
scales on the Conners CBRS, it is possible to get very Symptom scales.
extreme T-scores (i.e., > 100). To avoid over-interpretation
of these extreme values, all T-scores greater than 90 are The interpretation of discrepancies between DSM-IV-TR
reported as 90. Symptom Counts and T-scores has been updated to
reflect the changes in the T-score guidelines; see Table
In the original guidelines for interpreting T-scores, the 3. These discrepancies are to be expected, given that the
60 to 69 range was considered “elevated.” An additional Symptom Count and T-score are based on different metrics
condition, however, was imposed in that scores from 57 (i.e., absolute versus relative). Because the DSM-IV-TR
to 63 were specified as being in the “borderline” range. Symptom scale T-scores take age and gender into account,
The T-score interpretive guidelines have been adjusted so they may at times be more sensitive to atypicality for that
that the “borderline range” now reflects a specific category peer group, even if symptoms do not meet the absolute
of scores (i.e., scores that are between 1 and 1.5 standard symptom count level.
deviations above the mean; 60 to 64) and is described as
“High Average.”1 The standalone Conners Clinical Index™ (Conners CI™)
component items are identical to the Conners CI items that
The new interpretive guidelines are as follows: A T-score are included in the Conners CBRS. The interpretation of
in the “Very Elevated” range (i.e., > 2 standard deviations the Conners CI probability score does not differ between
above the mean) is very likely to indicate a significant area the two formats. The primary difference between the two
of concern. A T-score in the “Elevated” range (i.e., 1.5–2 formats is that the standalone form offers five subscale
standard deviations above the mean) usually indicates T-scores that are not available on the full-length Conners
significant concerns. A T-score in the “High Average” CBRS. Interpretation of these subscale T-scores should
range requires careful consideration and clinical judgment, follow the guidelines presented in Table 2 (i.e., the inter-
as this range is the borderline between typical and atypical pretive guidelines regarding T-score cut-offs also apply
levels of concern. T-scores falling below 60 generally to the Conners CI subscale T-scores).
indicate typical or absent concerns for the child’s age and
gender. (See Table 2 for a summary of these guidelines.)
Remember that clinical training and judgment are required
for responsible interpretation of any test score; these score
classification guidelines should not be applied automati-
cally without careful interpretation by a clinician.

Table 2. Understanding T-scores and Percentiles

T-score Percentile Guideline

70+ 98+ Very Elevated Score (Many more concerns than are typically reported)
65–69 93–97 Elevated Score (More concerns than are typically reported)
60–64 84–92 High Average Score (Slightly more concerns than are typically reported)
40–59 16–83 Average Score (Typical levels of concern)
< 40 < 16 Low Score (Fewer concerns than are typically reported)

1
Data analyses of the normative and clinical samples revealed that
2
scores in the 60 to 64 range are almost equally likely to occur This falls under Step 5 (Integrate Results) of the Step-by-Step
with clinical and non-clinical cases. Interpretation Guidelines in your Conners CBRS Manual.

2
 April, 2009

Table 3. Interpretation Guidelines for DSM-IV-TR Scores on the Conners CBRS

DSM-IV-TR DSM-IV-TR
Interpretation Guidelines
T-score Symptom Count
T-score and Symptom Count ≥ 65 At or above Significant features of the disorder are present in that setting.
are elevated DSM-IV-TR cut-off The symptoms are occurring in excess of what is typical for that
score youth’s age and gender.
This diagnosis should be given strong consideration.
Only T-score is elevated ≥ 65 Below DSM-IV-TR Features of this diagnosis are not prominent in that setting.
cut-off score The symptoms that are present are occurring in excess of what is
typical for that youth’s age and gender.
Although the current presentation is atypical for the youth’s age and
gender, there are not sufficient symptoms reported to meet DSM-
IV-TR symptomatic criteria for this disorder. The assessor may wish
to consider alternative explanations for why the T-scores could be
elevated in the absence of this diagnosis (e.g., another diagnosis
may be producing these types of concerns in that particular setting).
Only Symptom Count is ≤ 64 At or above Significant features of the disorder are present in that setting.
elevated DSM-IV-TR cut-off The symptoms are at (or below) developmental expectations for
score that age and gender.
Although the absolute DSM-IV-TR symptomatic criteria may
have been met, the current presentation is not atypical for this age
and gender. The assessor should carefully consider whether or not
symptoms are present in excess of developmental expectations (an
important requirement of DSM-IV-TR diagnosis).
T-score and Symptom Count ≤ 64 Below DSM-IV-TR Features of this diagnosis are not prominent in that setting.
are average or below cut-off score Any symptoms that are present are at (or below) developmental
expectations for that age and gender.
It is unlikely that the diagnosis is currently present (although
criteria may have been met in the past).

3. Defiant/Aggressive
Behaviors Scale
All three versions (Parent, Teacher, Self-Report) of the The underlying cause of an elevated Defiant/Aggres-
Conners CBRS include an Aggressive Behaviors scale. sive Behaviors score (i.e., defiant vs. aggressive) can
However, the true content of the scale measures not only be discerned by looking at the item-level responses
agression, but also defiance (behaviors that are consid- (Step 4 of the Step-by-Step Interpretation Guidelines
ered problematic but do not have overt expression of in your Conners CBRS Manual). The examination of
agression). To reflect this broader content, the scale name the responses to items in the scale can help determine if
has been changed from Aggressive Behaviors to Defiant/ the scale elevation is caused by defiant behaviors (e.g.,
Aggressive Behaviors. The characteristics of high scorers “Actively refuses to do what adults tell him/her to do,”
on the Defiant/Aggressive Behaviors scale are: “Argues with adults”), aggressive behaviors (e.g., “Physi-
cally hurts people,” “Is violent and aggressive towards
May be argumentative. May defy requests from others”), or both types of behaviors.
adults. May have poor control of anger and may
lose temper. May be physically and/or verbally
aggressive. May show violence, bullying, and
destructive tendencies. May seem uncaring. May
have legal problems.

3
 Conners Comprehensive Behavior Rating Scales™ Supplement

4. Violence Potential T-score, even if no other items are elevated). On the other
hand, it would take many of the less serious items to be
Indicator (VPI) endorsed before the T-score would be high.
Violence potential (i.e., the possibility that a youth may be
at risk for violent behavior) is a concern that has received
increased attention in recent years; as such, the Conners 5. Other Clinical Indicators
CBRS includes a Violence Potential scale. This scale When the Conners CBRS was released, for all Other
has been renamed Violence Potential Indicator (VPI) to Clinical Indicators (with the exception of the PDD items
better reflect the intended interpretation of the scale (i.e., on the Self-Report form), any response of 1, 2, or 3 would
elevated VPI scores do not necessarily indicate that the flag a recommendation that suggested the need for further
youth is/has been violent, rather, it is an indicator that the investigation. This approach was highly sensitive and
youth may behave violently in the future). The items on the ensured that any potential problems would be flagged.
VPI scale were informed by available literature on school However, for the Specific Phobia item, this approach has
violence and violence prevention. Items were selected to led to many non-clinical youth being flagged because it
represent concepts from the Warning Signs Antiviolence is very common for youth to be afraid of something at
Initiative of the American Psychological Association least occasionally. Therefore, the trigger point for this
(APA), which was developed in 1998 in an effort to help single item across all three forms (i.e., Parent, Teacher,
recognize the warning signs of teen violence.3 Other Self-Report) has been changed to 2 (i.e., ratings of 1 will
items were added based on findings from the Safe School no longer trigger a recommendation for follow-up). It
Initiative of the U.S. Secret Service and Department of should also be emphasized that for all of the Other Clinical
Education.4 Some of these items reflect extremely serious Indicators, these are flags only of potential problems and
existing violent and aggressive behaviors (e.g., “Uses a may not reflect a clinical condition.
weapon”); others are more benign (e.g., “Feels rejected”)
and are only indirectly related to the potential for future
violence.

Initially, like all other scales, all of the Violence Potential

scale items had the same “weight” (i.e., the raw score
is the sum of all item ratings). The scoring of the VPI
scale has been refined so that the more serious items now
“count” for more than the less serious items in computing
the raw score. Table 4 displays the scoring rules applied
to each item in the calculation of the VPI raw score. For
example, a rating of “1” on the relatively less severe item
“Annoys others” would be transposed such that only 0.5
was added to the raw score. Conversely, a rating of “1”
on a severe item, such as “Physically hurts others” would
be transposed such that 4 was added to the raw score.
These changes have affected the score calculations in
such a manner so that relatively few of the severe items
are required to trigger an elevated T-score (for example,
if “Uses a weapon,” and “Carries a weapon” are both
endorsed items, then the VPI scale may receive a high

3
American Psychological Association Practice Directorate.
(1998). Warning Signs of Teen Violence. Retrieved September
29, 2005, from http://helping.apa.org/featuredtopics/feature.php
4
Vossekuil, B., Fein, R., Reddy, M., Borum, R., & Modzeleski, W.
(2002). The Final Report and Findings of the Safe School Initia-
tive: Implications for the Prevention of School Attacks in the
United States. U.S. Department of Education, Office of Elemen-
tary and Secondary Education, Safe and Drug-Free Schools
Program and U.S. Secret Service, National Threat Assessment
Center, Washington, D.C. Retrieved July 5, 2007, from http://
www.secretservice.gov/ntac/ssi_final_report.pdf

4
 April, 2009

Table 4. VPI Transposing Rules

Item #
Transposing
Item Content
Parent Teacher Self-Report Rule

Trouble controlling anger 5 38 157

Feels rejected 21 198 n/a
Loses temper 45 3 58
Argues with adults 70 192 117 0→0
1 → 0.5
Poor school performance 130 61 18
2→1
Feels disrespected 139 153 41 3 → 1.5
Annoys others 163 128 134
Gets into trouble n/a n/a 7
Is lonely n/a n/a 11
Picked on or bullied 3 23 73
Disregard for others’ rights 13 57 120
Drinks alcohol 36 16 168
Lost interest/pleasure 53 46 93
Gets even 54 182 172
Starts fights 69 148 85
Seeks danger/risks/thrills 72 63 75
0→0
Angry/resentful 82 135 143 1→1
Uses drugs 131 75 68 2→3
3→5
Cruel to animals 161 115 112
Bullies/threatens/scares others 177 73 6
Damages/destroys others’ property 179 163 48
Gets high with household materials 182 29 141
Shows interest in weapons 170 196 n/a
Knows where to get a weapon n/a n/a 36
Gets into trouble at school 101 45 n/a
Made plans to hurt others 15 161 79
Carries a weapon 106 179 56
Uses a weapon 122 101 170 0→0
1→4
Gang membership 132 30 136
2→6
Violent/aggressive 140 190 4 3→8
Physically hurts others 144 52 144
Trouble with police 195 88 n/a
Note.  n/a = not applicable.

5
Conners Comprehensive Behavior Rating Scales™ Supplement

6
April, 2009

7
Conners Comprehensive Behavior Rating Scales™ Supplement