Bias and error rates for premorbid IQ estimators: comment on Veiel and Koopman (2001).

H. O. F. Veiel and R. F. Koopman (2001) advance statistical and legal theses. They correctly point out that the usual regression formula for estimating a pre-event IQ underestimates high IQs and overestimates low IQs (due to regression to the mean). They call this a conditional bias and show it can be sizeable. The author takes issue with their claim that a new estimator they propose should be used in place of the usual formulas, because it negates this statistical bias. Their argument against the usual estimator conflates statistical bias and legal bias. Their discussion in favor of their new estimator mentions, but does not derive a general formula for, a gross loss of precision entailed by use of the new estimator. The author quantifies this loss of precision and, using Veiel and Koopman's numerical example, shows that their estimator quadruples error.

Saccadic disinhibition in patients with acute and remitted schizophrenia and their first-degree biological relatives.

OBJECTIVE: Performance on measures of saccadic inhibition and control was investigated in a large family study of schizophrenia to evaluate the utility of using antisaccade task performance as an endophenotypic marker of genetic liability for schizophrenia. METHOD: Ninety-five patients with acute schizophrenia and 116 of their first-degree biological relatives, 13 schizophrenia patients whose illness was in full remission, 35 patients with acute psychotic affective disorder, and 109 nonpsychiatric comparison subjects were administered antisaccade and prosaccade tasks. RESULTS: Both schizophrenia patient groups had a greater number of errors on the antisaccade task than did the first-degree relatives and the affective disorder group, which both had more errors than the comparison subjects. Among the first-degree relatives of the probands with acute schizophrenia, relatives of poor-performing patients performed worse on the antisaccade task than relatives of patients with good performance. Reflexive errors were not likely the result of interfering psychotic symptoms, medication, or medication side effects. Although the schizophrenia patients demonstrated other signs of saccadic abnormalities, these problems, which were not observed in their relatives even though they had high antisaccade error rates, seem unlikely to account for the higher antisaccade error rate of the schizophrenia patients. CONCLUSIONS: These findings suggest that saccadic disinhibition is strongly associated with the genetic liability for schizophrenia.

Toward a resolution of the Rorschach controversy.

Comments are made about the articles comprising the first round of the Special Series on the Rorschach. G. Stricker and J. R. Gold (1999) and D. J. Viglione (1999) praised the Rorschach, but they consistently failed to cite negative findings. R. M. Dawes (1999) obtained results that provide modest support for the Rorschach, but one of his data sets is flawed. J. B. Hiller, R. Rosenthal, R. F. Bornstein, D. T. R. Berry, and S. Brunell-Neuleib (1999) reported the results of a meta-analysis, but, among other problems, their coders were not blind to the results of all the studies. J. Hunsley and J. M. Bailey (1999) made a strong case for concluding that there is no scientific basis for using the Rorschach. Recommendations are made for resolving the Rorschach controversy.

Clinical versus mechanical prediction: a meta-analysis.

The process of making judgments and decisions requires a method for combining data. To compare the accuracy of clinical and mechanical (formal, statistical) data-combination techniques, we performed a meta-analysis on studies of human health and behavior. On average, mechanical-prediction techniques were about 10% more accurate than clinical predictions. Depending on the specific analysis, mechanical prediction substantially outperformed clinical prediction in 33%-47% of studies examined. Although clinical predictions were often as accurate as mechanical predictions, in only a few studies (6%-16%) were they substantially more accurate. Superiority for mechanical-prediction techniques was consistent, regardless of the judgment task, type of judges, judges' amounts of experience, or the types of data being combined. Clinical predictions performed relatively less well when predictors included clinical interview data. These data indicate that mechanical predictions of human behaviors are equal or superior to clinical prediction methods for a wide range of circumstances.

Antisaccade performance in patients with schizophrenia and affective disorder.

The authors examined psychotic patients with schizophrenia, major depression, and bipolar disorder; "normal" participants; and 1st-degree relatives of patients with schizophrenia on an antisaccade task in which participants were instructed to move their eyes in the opposite direction of a target that moved unpredictably and abruptly either to the left or right of central fixation. Patients with schizophrenia were found to make significantly more errors than their relatives, and the latter made more errors than the controls. The poor performance of the relatives could not be attributed to their having a psychiatric disorder. Comparison of the 3 patient groups indicated that antisaccade deficits were more pronounced in schizophrenia and bipolar disorder.

The D5 dopamine receptor gene in schizophrenia: identification of a nonsense change and multiple missense changes but lack of association with disease.

To determine whether mutations in the D5 dopamine receptor gene (DRD5) are associated with schizophrenia, the gene was examined in 78 unrelated schizophrenic individuals (156 DRD5 alleles). After amplification by the polymerase chain reaction, products were examined by dideoxy fingerprinting (ddF), a screening method related to single strand conformational polymorphism analysis that detects essentially 100% of mutations. All samples with abnormal ddF patterns were sequenced. Nine different sequence changes were identified. Five of these were sequence changes that would result in protein alterations; of these, one was a nonsense change (C335X), one was a missense change in an amino acid conserved in all dopamine receptors (N351D), two were missense changes in amino acids that are identical in only some dopamine receptors and in only some species (A269V; S453C), and one was a missense change in a non-conserved amino acid (P330Q). To investigate whether the nonsense change (C335X), predicted to prematurely truncate the receptor protein and result in a 50% diminution of functional protein, was associated with schizophrenia, other neuropsychiatric diseases, or specific neuropsychological, psychophysiological, or personality traits, both case-control and family analyses were performed. No statistically-significant associations were detected with schizophrenia or other neuropsychiatric disease. There also were no significant associations between any one measure of neuropsychological function. However, a post-hoc analysis of combined measures of frontal lobe function hinted that heterozygotes for C335X may have a vulnerability to mild impairment, but these findings must be interpreted with caution.(ABSTRACT TRUNCATED AT 250 WORDS)

Maximizing heritability of a linear combination of traits.

In 1971 Jones proposed an approximate procedure for finding that linear combination of scores which has maximum heritability in a twin sample. I give an exact small-sample procedure. I point out two problems: such procedures can over-optimize the heritability by capitalizing on chance, and confidence intervals and significance tests are needed. I give an approach using James-Stein shrinkage estimation and bootstrapped standard errors to address these problems. It appears that confidence intervals may be quite broad. To reduce the width of the confidence intervals, one can accept some small-sample bias in exchange for smaller sampling errors. The James-Stein approach to estimating coefficients is used to achieve reduced confidence interval width. I illustrate with a computational example using personality data.

Comment on Levy et al. "Eye tracking dysfunction and schizophrenia".

We agree with Levy et al.'s conclusion (Schizophrenia Bulletin, Vol. 19, No. 3, 1993) that global quantitative measures of eye tracking dysfunction (ETD) have an important role in studies of ETD in schizophrenia. We clarify some misunderstandings of our own work contained in their review. In particular, an explanation is presented of how we computed a fit of our data to their Mendelian latent structure model, when testing for a fit to their hypothesis of essential genetic homogeneity. We also point out that our mixture analyses, as with our reported abnormality rates obtained by using cutting scores, are not in agreement with their reported rates of ETD in schizophrenia.

Simple regression-based procedures for taxometric investigations.

Certain theories of psychopathology postulate the existence of distinct latent populations of individuals. By analogy with biology, we call such latent populations "taxa" and we call the statistical testing of such theories "taxometrics." Taxometric procedures are robust; they do not make restrictive distributional assumptions. However, they have two disadvantages for nonstatistician users: (1) they are developed via algebra hard for many nonstatistician users intuitively to accept; and (2) computational software is not widely available. We address these problems by presenting a simple taxometric procedure, MAXSLOPE, based on regression plots for pairs of variables. This procedure is easily implemented using commonly available software and is intuitively rather easy to understand. We apply it to two artificial datasets. One dataset, used to explain the graphs, shows clear-cut evidence of taxa. The other example shows less clear grouping structure and is used to show that the proposed graphical procedure works even in nonideal cases. Comparisons are made with currently used procedures of cluster analysis and multivariate normal mixture analysis.

A latent trait finite mixture model for the analysis of rating agreement.

This article presents a latent distribution model for the analysis of agreement on dichotomous or ordered category ratings. The model includes parameters that characterize bias, category definitions, and measurement error for each rater or test. Parameter estimates can be used to evaluate rater performance and to improve classification or measurement with use of multiple ratings. A simple maximum likelihood estimation procedure is described. Two examples illustrate the approach. Although considered in the context of analyzing rater agreement, the model provides a general approach for mixture analysis using two or more ordered-caregory measures.

The association between lithium carbonate and smooth pursuit eye tracking among first-episode patients with psychotic affective disorders.

The association between treatment with lithium carbonate and smooth pursuit eye tracking performance was investigated in first-episode patients with psychotic affective disorders. The horizontal pursuit performance of patients with major depression and bipolar disorder who were receiving lithium carbonate was contrasted with that of patients not receiving lithium carbonate. In addition, the accuracy and quality of pursuit eye tracking was examined in bipolar patients whose lithium status changed from the time of initial testing to the time of retest 10 months later. For the combined group of depressed and bipolar patients, treatment with lithium carbonate was not associated with worse pursuit performance. Bipolar disordered patients on lithium did not differ in tracking proficiency from those not on lithium; bipolar patients whose lithium status changed from intake to retest also did not display a significant change in pursuit performance.

Smooth pursuit ocular motor dysfunction in schizophrenia: evidence for a major gene.

OBJECTIVE: Evidence suggests that poor eye tracking relates to genetically transmitted vulnerability for schizophrenia. The authors tested competing models for the genetic transmission of poor eye tracking in a search for major gene effects. METHOD: Samples from three studies (conducted in Minneapolis, New York, and Vancouver, B.C.) were pooled. Probands (N = 92) were diagnosed as schizophrenic by DSM-III criteria. Of the comparison subjects (N = 171), Vancouver patients were an epidemiologic first-episode group; at other sites unselected admitted patients were studied. First-degree relatives (N = 146) of 65 probands were also studied. Eye tracking was measured while subjects followed a horizontally moving, sinusoidally driven (0.4 Hz) spot of light on a screen. Performance was quantified by root mean square error. Data analysis was by complex segregation analysis (Bonney's class D regressive models). RESULTS: A single major gene is needed to account for poor eye tracking in schizophrenic patients and their relatives. This gene alone can explain about two-thirds of the variance in eye tracking performance. A single gene alone (regardless of dominance) will, however, not account for the data; polygenic factors are also required. CONCLUSIONS: Results support postulation of a single gene for ocular motor dysfunction, which may be a risk factor for schizophrenia. Eye tracking may be useful as a gene carrier test in genetic studies of schizophrenia.

Cognitive therapy and pharmacotherapy for depression. Singly and in combination.

Cognitive therapy and imipramine hydrochloride tricyclic pharmacotherapy, each singly and in combination, were compared in the treatment of nonpsychotic, nonbipolar depressed outpatients. One hundred seven patients were randomly assigned to 12 weeks of active treatment; 64 patients completed the full course of treatment. Rates of attrition were high but not differential. Cognitive therapy and pharmacotherapy did not differ in terms of symptomatic response, either in the primary analyses or in secondary analyses restricted to more severely depressed outpatients. Initial severity did predict response within pharmacotherapy alone, but not within cognitive therapy. Combining cognitive therapy with pharmacotherapy did not markedly improve response over that observed for either modality alone, although such nonsignificant differences as were evident did favor the combined treatment. Two patients died as a consequence of suicide attempts, both of which involved study medication.

Differential relapse following cognitive therapy and pharmacotherapy for depression.

Patients successfully treated during a 3-month period with either imipramine hydrochloride pharmacotherapy, cognitive therapy, or combined cognitive-pharmacotherapy were monitored during a 2-year posttreatment follow-up period. Half of the patients treated with pharmacotherapy alone continued to receive study medications for the first year of the follow-up. All other patients discontinued treatment at the end of the acute treatment phase. Patients treated with cognitive therapy (either alone or in combination with medication) evidenced less than half the rate of relapse shown by patients in the medication--no continuation condition, and their rate did not differ from that of patients provided with continuation medication. It appears that providing cognitive therapy during acute treatment prevents relapse. Whether this preventive effect extends to recurrence remains to be determined.

Smooth-pursuit eye movement dysfunction and liability for schizophrenia: implications for genetic modeling.

Forty-one nonpsychiatric subjects, 38 probands with schizophrenia, and 99 of their relatives were studied. Oculomotor functioning was bimodally distributed for probands and relatives. Oculomotor dysfunction was not present in all families with a schizophrenic proband. In those families in which it was present, there were significant phenotypic correlations between oculomotor functioning and schizophrenia-related characteristics. The patterns of familial resemblance in the families in whom oculomotor dysfunction was present were consistent with nonadditive genetic variance contributing both to oculomotor dysfunction and to the relationship between oculomotor dysfunction and clinical symptoms. These results suggest that schizophrenia may be etiologically heterogeneous and that oculomotor dysfunction may help to identify nonadditive genetic variance for this disorder.

Follow-up and family study of anxious depression.

OBJECTIVE: The failure of the concept of anxious depression to find its way into DSM-III-R led the authors to conclude that a further report on the occurrence of anxiety symptoms in depressed subjects is indicated. METHOD: The subjects were 327 consecutively evaluated inpatients and outpatients with primary unipolar depressive disorder at five university medical centers participating in the National Institute of Mental Health Collaborative Program on the Psychobiology of Depression--Clinical Studies. The authors restricted their sample selection to patients with primary depressive disorder so that patients with other preexisting psychiatric disorders, especially anxiety disorders, would not contaminate the symptom picture, family studies, or follow-up. The examined six anxiety symptoms and derived a new anxiety summary score to show the effect of anxiety in depression on family data and 5-year outcome. RESULTS: Depressed subjects with higher ratings for anxiety took longer to recover. There was also a significant relationship between anxiety in depressed probands and the risk for primary unipolar depressive disorder, but not anxiety disorders or alcoholism, among 832 blindly interviewed first-degree relatives. CONCLUSIONS: These data confirm the usefulness of subdividing depressed patients according to anxiety symptoms: psychic and somatic symptoms of anxiety, taken together, significantly predict family illness and course. The data also emphasize the wisdom of requiring that generalized anxiety disorder not be diagnosed in the presence of a mood disorder. Clearly, symptoms of anxiety coexist with depression and need to be recognized for the effective treatment of the underlying depressive disorder.

Psychometric detection of schizotypy: perceptual aberration and physical anhedonia in relatives of schizophrenics.

We administered scales of Perceptual Aberration (PERAB) and Physical Anhedonia (PHYSAN), traits that may be related to risk for schizophrenia, to 54 schizophrenics, 146 of their first-degree relatives (evaluated for schizophrenia-related disorders), and 178 normal subjects (screened for psychotic disorders in them or their relatives). For both scales, there was a significant effect of group membership. For the PERAB scale, the schizophrenics had higher scores than the normal subjects, who had higher scores than the relatives. For the PHYSAN scale, schizophrenics had higher scores than their relatives, who had higher scores than the normal subjects. Patterns of familial correlations also suggested that physical anhedonia, but not perceptual aberration, may be familial among schizophrenics and their relatives. The PHYSAN scale, but not the PERAB one, may be a useful indicator of liability for schizophrenia among the relatives of affected probands.

When is a diagnosis worth making? A statistical comparison of two prediction strategies.

For a setting in which the sole goal is to predict a criterion accurately, two strategies are compared, (1) multiple linear regression directly from a set of predictors and (2) using predictors to diagnose individuals and then using only the diagnosis to predict the criterion. These are mathematical models of two common methods for making clinical predictions. This article derives equations for each statistical strategy's validity (and a formula comparing them, for a special case). Prediction accuracies are compared over a simplified but broad parameter space and four conclusions follow. Changing disorder prevalences (in the range 0.1 to 0.5) have small effect on the relative preferability of the two strategies; the effect of varying prevalence differs according to population separation on predictors and criterion. As within-population covariances of predictors with criterion decline below zero (assuming variables are scaled so that the less common population scores higher on predictors and criterion), diagnoses become increasingly preferable as a prediction strategy; as they rise above zero the opposite trend is observed. As populations become better separated on predictors or criterion, diagnoses compare more favorably. Most importantly, over almost all of the parameter space which one would expect to encounter in clinical psychology and psychiatry, multiple linear regression is much superior.