the role of attention in affect perception: an examination of mir

The Role of Attention in Affect Perception:An Examination of Mir sky's Four Factor Model

of Attention in Chronic Schizophrenia

by Dennis R. Combs and Wm. Drew Qouvier

Abstract

Attentional skills among people with schizophreniamay be related to deficits in affect perception. Suchdeficits can dramatically inhibit appropriate socialfunctioning. We examined attention and affect percep-tion in a sample of 65 people diagnosed with chronicschizophrenia. We used Mirsky's four factor model ofattention to assess attentional functioning. To measureaffect perception, we used two reliable measures ofemotion recognition, the Bell-Lysaker EmotionRecognition Test and the Face Emotion IdentificationTest. Multiple regression analysis showed that all fourattentional factors and a diagnosis of paranoid schizo-phrenia were significantly predictive of affect percep-tion scores. In contrast, psychiatric symptoms, medica-tion levels, demographic variables, verbal fluency, andface recognition scores were not predictive of affectperception scores. The four factors of attentionaccounted for 78 percent of the variance in affect per-ception scores. These results emphasize the role thatattentional abilities play in affect perception for peoplewith schizophrenia.

Keywords: Attention, affect perception, schizo-phrenia, social functioning.

Schizophrenia Bulletin, 30(4):727-738, 2004.

Many researchers believe that schizophrenia is essentiallya social-cognitive disorder (Bellack 1992; Penn andMueser 1996; Penn et al. 1997; Kohler et al. 2000).People diagnosed with schizophrenia demonstrate varioussocial impairments such as poor affect perception, defi-cient social skills, decreased recognition of nonverbalcues, and social competence deficits (Bellack 1992; Salemet al. 1996; Penn et al. 1997). Schizophrenia has also beenassociated with a range of information-processing andneuropsychological deficits (Saykin et al. 1991, 1994;Schwartz et al. 1992; Zalewski et al. 1998), and someresearchers believe that problems in cognitive functioningmay affect people's ability to learn, exhibit, and express

social skills and behaviors (Morrison et al. 1988ft; Bellack1992; Penn et al. 1997, 2000). Researchers have becomeincreasingly interested in the possible link between infor-mation processing and social functioning (Bellack 1992;Cornblatt et al. 1992; Freedman et al. 1998; Bellack et al.1999; Penn et al. 2000; Addington et al. 2001). However,studies examining this relationship are few and have beeninconsistent in their findings regarding which cognitivevariables are linked to adaptive social functioning (Green1996).

One aspect of social behavior needing further study isthe ability to perceive and identify another person's affec-tive or emotional state (affect perception). Many empiri-cal studies have shown consistent deficits in affect percep-tion among those diagnosed with schizophrenia(Doughtery et al. 1974; Muzekari and Bates 1977; Novicet al. 1984; Cramer et al. 1989; Archer et al. 1992; Kerrand Neale 1993; Schneider et al. 1995; Salem et al. 1996;Mandal et al. 1998; Penn and Combs 2000; Penn et al.2000), and some researchers believe that these deficits arethe most debilitating of all the social impairments foundin schizophrenia (Morrison et al. 1988ft) These findingsraise the question of what variables contribute to thesedeficits.

Deficits in affect perception and other social behav-iors may stem from cognitive impairments associatedwith schizophrenia (Bellack 1992; Green 1993;Addington and Addington 1998; Mandal et al. 1998;Morrison et al. 1988ft). Bellack (1992) stated that success-fully completing most social perception tasks requiressubstantial cognitive demands. A better understanding ofthe relationship between information-processing abilitiesand affect perception may lead to a better understandingof the characteristics of the disorder. In fact, several stud-ies have found that affect-perception skills were related toadaptive social functioning and social skills in both inpa-

Send reprint requests to Dr. D.R. Combs, University of Tulsa, LortonHall, Room 303, 600 S. College, Tulsa, OK 74104; e-mail: [email protected].

727Downloaded from https://academic.oup.com/schizophreniabulletin/article-abstract/30/4/727/1930996by gueston 16 February 2018

Schizophrenia Bulletin, Vol. 30, No. 4, 2004 D.R. Combs and Wm. D. Gouvier

tient and outpatient settings (Penn et al. 1996; Mueser etal. 1996; Ihnen et al. 1998).

Emerging evidence indicates that a particular cogni-tive variable, attention, may be an important componentof affect perception. Theoretically, deficient attentionalskills could impair affect perception by interfering withthe ability to attend to and decode facial stimuli (Bellack1992; Addington and Addington 1998; Mandal et al.1998; Morrison et al. 1988fc). Bruce and Young (1986)posited an important role for visual attention in facerecognition and in affect/expression analysis. This theorywas empirically demonstrated in a series of studies byPhillips and David (1997, 1998) in which people withdelusions scanned more nonrelevant areas of the face.Thus, affect perception requires that people select whichparts of the face to attend to and then maintain their atten-tion to collect important information about another's emo-tional state (Green 1996; Bryson et al. 1997; Addingtonand Addington 1998; Morrison et al. 1988a). Quite sim-ply, if a person cannot fully attend to facial stimuli, thenhis or her capacity to decode and interpret emotionalexpressions will be concomitantly impaired.

Several empirical studies have examined the role ofneuropsychological variables in affect perception. Anexamination of these studies showed consistent evidencethat attention was a key variable in affect perception(Bryson et al. 1997; Addington and Addington 1998; Keeet al. 1998; Kohler et al. 2000)—arguably more importantthan memory, learning, motor functioning, and reasoning.Furthermore, because attention is a necessary precursorfor processing incoming information, attention problemscan affect many other higher order cognitive abilities(Mapou 1995).

Although attention may be a crucial factor in affectperception (Archer et al. 1992; Kerr and Neale 1993;Bryson et al. 1997; Mandal et al. 1998), only a few stud-ies have specifically examined the aforementioned linkbetween attention and affect perception. Bryson et al.(1997), Addington and Addington (1998), Kee et al.(1998), and Kohler et al. (2001) have all found some sup-port for the role of attention in affect perception, althoughnot all studies agree on its importance (Schneider et al.1995; Morrison et al. 1988a). Specifically, attention span,the ability to shift attention, and sustained attention haveall been found to be related to affect perception scores.However, these studies used unvalidated attentional mea-sures, and these measures were not selected based on cur-rent theoretical models of attentional functioning (i.e.,Mirskyetal. 1991).

In sum, theoretical and research evidence suggeststhat attention may be important in affect perception(Morrison et al. 1988b; Bellack 1992; Bryson et al. 1997;Mandal et al. 1998). To identify different emotional states,

facial information must be attended to and perceived(Bellack 1992). Various research studies on affect percep-tion have stated that attentional problems may be respon-sible for the observed deficits, but only a few haveexplored this topic empirically. Attention should beassessed in a way that is consistent with research on themultifactorial nature of the attentional process (Solhbergand Mateer 1989; Mirsky et al. 1991). Finally, the rela-tionship between attention and affect perception couldhave implications for the cognitive rehabilitation of thesedeficits (Hogarty and Flesher 1992; Green 1996). Thestudy of attention in affect perception therefore hasimportant theoretical, empirical, and clinical merit.

The purpose of this study was to explore the relation-ship between affect perception and attention in a sampleof people with chronic schizophrenia using Mirsky's fourfactor model of attention. Mirsky et al. (1991) identifiedfour factors of attention (Shift, Sustain, Encode, andFocus-Execute) based on a Principal ComponentsAnalysis (PCA) of common neuropsychological mea-sures. The four factor model of attention has been repli-cated in other samples of people with chronic schizophre-nia (Steinhauer et al. 1991; Kremen et al. 1992). Previousresearch findings suggest that the Sustain, Shift, andEncode factors will best predict affect perception scores.In addition, because there is evidence that people withparanoid schizophrenia show improved affect perceptionscores, the relationship between diagnostic subtype andaffect perception was also examined (Kline et al. 1992;Lewis and Garver 1995). Finally, two other neurocogni-tive measures (verbal fluency and general face perception)were included to examine what role, if any, these skillsplay in affect perception.

Method

Participants. The participants were 65 people diagnosedwith chronic schizophrenia who were recruited from twolarge State psychiatric hospital settings in Louisiana:Southeast Louisiana State Hospital and the EasternLouisiana Mental Health System. Both settings are ter-tiary care treatment centers for people with chronic men-tal illness. Table 1 summarizes participant demographicand clinical characteristics. A total of 36 males and 29females participated in the study. By ethnicity, 17 wereCaucasian, 47 were African-American, and 1 was Asian-American. There was no difference in the gender and eth-nic makeup of the total sample, x2 (2) = 1.3, p = 0.50.Comparison tests did not reveal any differences in atten-tional or affect perception measures based on gender, eth-nicity, or hospital site. To be eligible for the study, partici-pants were required to have a DSM-IV diagnosis of


The Role of Attention in Affect Perception Schizophrenia Bulletin, Vol. 30, No. 4, 2004

Table 1. Summary of participant demographics

Variable

Age (yrs)

Education (yrs)

Gender (n)MaleFemale

Ethnicity (n)CaucasianAfrican-AmericanOther

SCID-I/P diagnosis (n)Paranoid schizophreniaUndifferentiated schizophreniaDisorganized schizophreniaSchizoaffective disorder

CPZ equivalents (mg/d)1

Antipsychotic medication type (n)TypicalAtypicalCombination

Anticholingeric medication (%)

BPRS total score

AIMS score

Length of illness (mos)

Length of stay (wks)

Hospital admits (n)

WRAT-III, reading

Total samplemean (SD)

40.7 (7.9)

11.2(1.7)

3629

1747

1

4084

13

854.0 (494.5)

15341636

53.8 (5.9)

0.17(0.57)

218.8(97.7)

86.0(110.7)

11.1 (17.3)

79.8 (9.0)

EasternLouisiana Mental

Health Systemmean (SD)

41.0(8.0)

11.1 (1.8)

2419

1032

1

2562

10

864.2 (534.7)

105181544

54.3 (6.0)

0.21 (0.64)

220.5(102.9)

84.4(123.4)

13.6(20.8)

79.4 (8.6)

SoutheastLouisiana State

Hospitalmean (SD)

40.2 (8.0)

11.5(1.6)

1210

7150

15223

834.0(415.8)

161

22

52.8 (5.8)

0.09 (0.43)

216.5(88.9)

89.2 (82.9)

6.4(4.1)

80.7(10.0)

Note.—AIMS = Abnormal Involuntary Movement Scale; BPRS = Brief Psychiatric Rating Scale; CPZ = chlorpromazine; SCID-I/P =Structured Clinical Interview for DSM-IV; WRAT-III = Wide Range Achievement Test III.1 Medication dosages were converted to chlorpromazine equivalents.

schizophrenia, schizoaffective disorder, or psychotic dis-order not otherwise specified. All participants wereassessed using the Structured Clinical Interview for theDSM-IV (SCID-I/P; First et al. 1995) to confirm diagno-sis of schizophrenia. To examine the effects of diagnosticsubtype on affect perception, two groups were formed: Aparanoid group (paranoid schizophrenia) and a nonpara-noid group (disorganized, undifferentiated, and schizoaf-fective). Psychiatric symptomatology was assessed usingthe Brief Psychiatric Rating Scale (BPRS; Lukoff et al.1986). Tardive dyskinesia symptoms were assessed by areview of Abnormal Involuntary Movement Scale (AIMS)scores conducted by hospital staff. Data on type and

dosage of antipsychotic medication were collected, andmedication dosages were converted into standardizedchlorpromazine equivalents (see Lehman and Steinwachs1998 and Bezchlibnyk-Butler and Jeffries 2002 for CPZconversion rates).

Participants who had documented substance abuseproblems were required to be in remission or in a con-trolled environment for 3 months before participating.Substance abuse problems were ruled out by a review ofsubstance screening tests on admission and by theSCID-I/P substance abuse module. Other exclusion crite-ria included having a documented neurological condition(e.g., stroke, traumatic brain injury, or seizure disorder),



being too disorganized to give consent or complete theprotocol, being deemed unsuitable for inclusion by hospi-tal staff, or reading level below the fourth-grade level asassessed by the Wide Range Achievement Test III(WRAT-III) reading test (Wilkinson 1993).

Procedure for Selection of Participants. Participantswere selected using a methodology employed in previousresearch studies in these settings (Penn and Combs 2000;Penn et al. 2000). Potential participants first received ageneral overview of the study. Next, a list of all peoplewith diagnoses that met study eligibility criteria was con-structed from the hospital roster. Participants either volun-teered to take part in the study after the general overviewor were identified (based on diagnosis) and approachedfor participation. One person refused to participate in thestudy, and two people refused to continue testing. In allthree cases, the subject became disorganized, psychotic,and confused.

Demographic and Clinical Measures. A basic demo-graphic questionnaire was used to obtain backgroundinformation for each participant (e.g., length of illness,duration of current stay, number of inpatient hospitaliza-tions, AIMS score, and medication type and dosage). Thisinformation was obtained from hospital records and aninterview with the participant.

The SCID-I/P was used to derive a clinical psychiatricdiagnosis based on the DSM-IV system (First et al. 1995).The BPRS assessed each participant's current level ofsymptomatology over the previous 1- to 2-week period(Lukoff et al. 1986). The BPRS contains 24 items (ratedon a scale of 1 to 7) that cover a range of psychiatricsymptoms. The BPRS can be broken down into four fac-tor scores: anergia, affect, thought disorder, and disorgani-zation (Mueser et al. 1997). The principal investigatorwas reliability trained on the BPRS and SCID-I/P (Pennand Combs 2000; Penn et al. 2000). The WRAT-III(Wilkinson 1993) reading subtest was used to screen par-ticipants for problems in reading.

Measures of Attention. The measures of attention usedin this study were selected based on Mirsky's four factormodel of attention: Focus-Execute, Encode, Sustain, andShift (Mirsky et al. 1991, 1995). The Shift factor wascomposed of scores from the Wisconsin Card Sorting Test(WCST; Heaton 1981). WCST variables of interest werepercentage correct, number of errors, and categories com-pleted. The person must be able to disengage or shiftattention from the previous response to the new set (e.g.,color to form), which may be different from the reason-ing/problem-solving aspects of the test (Mirsky et al.

1995; Spreen and Strauss 1998). A computerized versionof the WCST was used for ease of examination. TheSustain factor was based on scores from the conventionalversion of the Continuous Performance Test (CPT;obtained from the University of California at LosAngeles), which required the person to respond by click-ing a mouse button each time the number "0" appeared(Nuechterlein 1991). Of the 480 total numbers presented,only 120 were targets. Variables of interest from the CPTincluded number of hits, total number of errors (missesand false alarms), and mean reaction time (Mirsky et al.1991, 1995; Kremen et al. 1992). The CPT also provideda sensitivity index (d'; based on signal detection theory),which measured the person's sensitivity to correctlyrespond to targets and ignore distracter numbers. TheFocus-Execute factor was based on scores from the TrailMaking Test (Reitan and Davison 1974; time in seconds)and Digit Symbol-Coding subtest (number correct) fromthe WAIS-III. The final factor, Encode, was based onscores from the Digit-Span subtest (number correct) andArithmetic subtest (number correct) of the WAIS-III.

Affect Perception MeasuresBell-Lysaker Emotion Recognition Test (BLERT).

The BLERT is a 21-item videotaped presentation of sevendifferent emotional states: happiness, sadness, anger, fear,disgust, surprise, and no emotion (Bell et al. 1997; Brysonet al. 1997). Each emotional state was presented for 10seconds, and the subject had to decide which affectivestate was presented. Each emotion was displayed by amale actor who recited a series of three standard mono-logues concerning his job. The BLERT total score (range0 to 21) was used in the analyses. The BLERT has goodcategorical stability data (K = 0.76) and test-retest stability(5-month test-retest reliability was 0.76). The BLERT alsohas good discriminant and convergent validity (Bell et al.1997; Bryson et al. 1997). In the present study, the inter-nal consistency for this scale was 0.79.

Face Emotion Identification Test (FEIT). TheFEIT was developed by Kerr and Neale (1993) using stillphotograph pictures taken from the work of Ekman(1976) and Izard (1971) and based on the generalizeddeficit model (Chapman and Chapman 1973, 1978). TheFEIT consists of 19 (score range 0 to 19) videotaped pic-tures of six different emotional states: happiness, sadness,anger, surprised, afraid, and ashamed. The FEIT wasdeveloped as an affect perception test with acceptablereliability and validity. Previous reliability results showedan internal consistency value ranging from 0.56 to 0.71(Kerr and Neale 1993). Comparable reliability resultswere replicated in another study using a similar state hos-pital sample (Penn et al. 2000). The FEIT has demon-strated good discriminant and convergent validity (Kerr



and Neale 1993). In the present study, the internal consis-tency for this scale was 0.75.

Other Neurocognitive Measures. To evaluate the person'sverbal fluency, the Controlled Oral Word Association test(COWAt; Benton et al. 1994; Spreen and Strauss 1998) wasadministered. The test evaluates whether deficits in emotionrecognition might be due to problems in general verbal flu-ency (language production) that might prevent the subjectfrom generating the names for different emotions (Chen etal. 2000). The COWA requires the person to verbally gener-ate as many items beginning with the letters F, A, and S aspossible in 60 seconds, and a total score across all three let-ters was computed. The short form of the Benton Test ofFacial Recognition (TFR; Benton et al. 1983) was adminis-tered as a control measure for the affect perception tasks.The test presented 27 pictures of different faces that the per-son matched in identity to a target face. Because the testcontrolled for the emotional content of the faces, it couldserve as a pure face recognition test. The TFR has been usedin previous studies to compare face perception and affectrecognition (Salem et al. 1996; Penn et al. 2000).

General Procedures. The main component of the studywas the administration of the measures of attention andaffect perception. The measures of attention were admin-istered individually and randomized before administra-tion. The CPT and WCST were administered using a com-puterized testing format. These were followed by theCOWAt and the TFR. Finally, the affect perception mea-sures (BLERT and FEIT) were administered. Each partici-pant was compensated $5 for time and effort after com-pleting the entire study protocol. The research protocoltook approximately 1.5 to 2 hours to complete, and amplerest periods and breaks were provided to reduce the stressand fatigue associated with intensive testing.

Data Analytic Plan. The statistical analysis begins withsummary scores based on the attentional and cognitivemeasures, followed by correlations between affect per-ception and the demographic and clinical variables.Because Mirsky's four factor model has been replicatedpreviously, and because the attentional factors found inthis study (Shift, Sustain, Encode, and Focus-Execute)were very similar to Mirsky's, the derivation of the fac-tors using PCA is not reported. The factor scores com-puted from the PCA were used a predictor variables inthe subsequent regression analyses. The dependent vari-able of interest is a standardized affect perception scorecomposed of scores from the BLERT and FEIT. To deter-mine which attentional factors were related to affect per-ception scores, we conducted a stepwise multiple regres-sion analysis.

Results

Summary Scores and Correlations. Summary scoresfor the attentional, affect perception, and cognitive mea-sures are presented in table 2. These scores are reported asraw scores (Mirsky et al. 1991). In addition, correlationsbetween affect perception and the demographic and clini-cal variables can be found in table 3. Variables signifi-cantly correlated with affect perception included theCOWAT (r = 0.326, p = 0.002), the TFR (r = 0.450, p =0.0001), BPRS total score (r = -0.329, p = 0.008), andpsychiatric subtype diagnosis (r = -0.295, p = 0.008).

Multiple Regression Analysis. We conducted a stepwisemultiple regression analysis to determine which factor ofattention was most predictive of affect perception scores.Predictor variables were the four factor scores generatedfrom the PCA and other study variables related to affectperception scores (table 3). The BPRS total score,COWAt score, TFR score, and diagnostic subtype (para-

Table 2. Summary scores on attentional,neurocognitive, and affect perception measures1

Variable

Trails A (sec)

Trails B (sec)

Digit Symbol Coding

Arithmetic

Digit Span

CPTHits (n)Errors (n)

Reaction time (ms)<f index score

WCSTCorrect (%)Errors (n)Categories completed

COWAt

TFR (0-54)

BLERT (0-21)

FEIT (0-19)

Mean (SD)

60.9 (21.4)

144.2(57.9)

33.6(13.3)

9.7 (2.2)

12.3(2.5)

101.0(28.2)30.9 (40.2)

455.6(132.7)3.6(1.5)

59.4 ( - )49.7(21.0)

3.4 (2.0)

25.4 (4.5)

43.9 (4.5)

12.1 (4.4)

11.2(3.7)

Note.—BLERT = Bell-Lysaker Emotion Recognition Test; COWAt =Controlled Oral Word Association test; CPT = ContinuousPerformance Test; FEIT = Facial Emotion Identification Test; TFR =Test of Facial Recognition; WCST = Wisconsin Card Sorting Test.1 All scores are reported as raw scores.



Table 3. Correlations between demographic,clinical, and affect perception variables

Variable

Age (yrs)Education

Facial recognition

Verbal fluency

Diagnosis1

BPRS total

Length of illness(mos)

Length of stay(wks)

Hospitalizations (n)

AIMS

Medication dose(CPZ equivalent)

Medication type2

Affect perception score(correlation coefficient)

-0.200.24

0.45*

0.37*

-0.29*

-0.32*

-0.20

-0.15

-0.24

-0.21

-0.02

0.03

Note.—AIMS = Abnormal Involuntary Movement Scale BPRS =Brief Psychiatric Rating Scale; CPZ = chlorpromazine.1 Diagnosis was dummy coded as paranoid versus nonparanoid.

dedication type was coded as typical versus atypical.*p<0.05

noid versus nonparanoid) were included as additional pre-dictor variables. Note that when people with schizoaffec-tive disorder (n - 13) were removed from the diagnosisvariable, the results were unchanged.

A composite affect perception score (based onBLERT and FEIT) was the dependent variable in thisanalysis. This composite total score represented a moreglobal, comprehensive measure of affect perception.Because scores from the BLERT and FEIT were found tobe highly correlated (r - 0.85, p = 0.0001), the two affectperception scores were converted into standardized zscores based on sample means, and a composite affectperception score was computed (mean z scores fromBLERT and FEIT). The conversion of these scores into zscores was needed to make the two measures comparable(see Salem et al. 1996 for a similar procedure) becausethe BLERT and FEIT have different numbers of test itemsand have different presentation formats (the BLERT hasan audiovisual format and the FEIT is visual only). Nodifferences in the regression results existed when theFEIT and BLERT were analyzed separately. The results ofthe stepwise multiple regression analysis are presented intable 4.

Overall, the four factors of attention and psychiatricdiagnosis accounted for 81 percent of the variance inaffect perception scores. The four factors of attentionaccounted for 78 percent of the variance in affect percep-tion scores and diagnostic subtype accounted for 3 percentof the variance. The Shift factor had the highest predictivevalue (34.6%) followed by Encode (18.5%), Focus-Execute (14.7%), and Sustain (9.7%) factors. Examiningthe effect of diagnosis revealed that participants withparanoid schizophrenia had higher affect perceptionscores than participants with nonparanoid schizophrenia,F (1,63) = 7.3, p - 0.009. There were no differences inBPRS total or factor scores between the paranoid andnonparanoid groups. The other variables (COWAT, TFR,and BPRS) were not included in the stepwise analysis.1

Supplementary Analysis. To determine whether atten-tion was also important in face recognition, the four atten-tional factors were regressed against scores from the TFR.The four attentional factors predicted 20 percent of thevariance in TFR scores (r - 0.45, r2 = 0.20) comparedwith 78 percent in the affect perception scores. TheFocus-Execute factor was the only significant predictorvariable (r[64] = -2.0, p = 0.04).

Discussion

The stepwise multiple regression analysis showed that allfour factors of attention, along with diagnostic subtype,were significantly related to affect perception scores. TheShift and Encode factors were found to be the strongestpredictors of affect perception performance. The Focus-Execute and Sustain factors were also significant predic-tors but accounted for less variance. The four attentionalfactors accounted for 78 percent of the variance in affectperception scores. The attentional variables predictedemotion perception scores beyond that accounted for bygeneral face perception, which suggests that basic faceperception was not a significant factor for the affect per-ception task. In addition, the lack of a relationship withverbal fluency suggests that problems with affect percep-tion are not related to verbal fluency deficits. No effect onaffect perception was found for the BPRS symptomsscores, which will be discussed in more detail below.Furthermore, the attentional factors were found to be min-

1 A direct regression analysis was conducted to examine the problemsassociated with stepwise regression methods because of the eliminationof important variables, which share variance with the selected predictors(Thompson 1995). BPRS, TFR, and COWA were still not found to besignificant predictors and did not account for any appreciable variance inaffect perception scores.



Table 4. Stepwise Regression Analysis Results1

Step/variable

1.2.3.4.5.

ShiftEncodeFocus-ExecuteSustainDiagnosis

r

0.5970.7360.8300.8870.905

0.3560.5410.6890.7860.819

A/2

0.3460.1850.1470.0970.030

P0.5810.400

-0.3930.303

-0.186

f

10.4"7.0"

-7 .0"5.4"

-3.3*

Zero order

0.5970.430

-0.3840.312

-0.2951 Excluded variables in this analysis were the Brief Psychiatric Rating Scale total score, Controlled Oral Word Association Test, and theBenton Test of Facial Recognition; Diagnosis was dummy coded as paranoid versus nonparanoid.

' p< 0.005; **p< 0.001

imally predictive of face recognition scores (supplemen-tary analysis), which suggests that attention may be moreimportant in affect perception and less important for gen-eral face recognition (Kohler et al. 2001). The Focus-Execute factor was the only attentional variable predictiveof face recognition. This result may be due to thedemands of the test, in which the subject must select(focus) the target face and then respond (execute).

Bryson et al. (1997) and Kee et al. (1998) presentedsimilar evidence that encoding skills, sustained attention,and the ability to shift attention were related to affect per-ception scores. Note that in the Bryson et al. study (1997),cognitive-attentional variables accounted for only 34 per-cent of the variance in affect perception as measured bythe BLERT. The attentional variables used in the presentstudy accounted for much more variance (78% versus34%). The finding that participants with paranoid schizo-phrenia showed better affect perception scores is consis-tent with the results of Kline et al. (1992), Lewis andGarver (1995), and Toomey et al. (2002). Magaro (1981)and Strauss (1993) argued that people with paranoidschizophrenia have more intact cognitive abilities thanpeople with nonparanoid schizophrenia. Because no dif-ferences in psychiatric symptoms on the BPRS existedbetween the paranoid and nonparanoid participants, wespeculate that cognitive factors (which were not directlyassessed) may be a source of the differences found in thisstudy.

Surprisingly, no empirical relationship existedbetween affect perception and current psychiatric symp-toms (positive and negative), medication type or dosage,demographic, or other illness-related variables (table 3),suggesting that the affect perception scores found in thisstudy may not be an artifact of these variables. Regardingillness-related variables, one study showed that a higherthe number of hospitalizations was paradoxically associ-ated with better FEIT scores (Salem et al. 1996), whereasanother study showed that the length of the current treat-ment episode was inversely correlated with the FEIT(Mueser et al. 1996). This present study did not find any

association between length of illness or length of currenttreatment episode and affect perception scores. Previousresearch studies showed a modest association betweenaffect perception and psychiatric symptoms, with negativesymptoms associated with lower affect perception scoresand more visual-spatial processing deficits (Borod et al.1993; Kohler et al. 2000; Penn et al. 2000; Strauss 1993;Schneider et al. 1995). Although the BPRS total score wasinitially correlated with affect perception scores in thepresent study, its influence was better accounted for by theattentional and diagnostic variables. Researchers havehypothesized that the link between negative symptomsand poor affect perception stems from lower social func-tioning and poorer premorbid adjustment or the presenceof structural brain impairments. However, not all studieshave shown an association between negative symptomsand affect perception scores (Salem et al. 1996; Toomeyet al. 2002), and currently the role of symptoms in affectperception remains unclear. Finally, and somewhat sur-prisingly, no relationship between affect perception andmedication type or dosage was found. Many believe thatthe newer atypical antipsychotic medications mayimprove cognitive abilities (Nagamoto et al. 1996;Meltzer and McGurk 1999), which may in turn improveaffect perception (Green 1996; Stip and Lussier 1996).However, research on this topic is limited (Toomey et al.2002), and it is not known if the improvement in affectperception is due to indirect enhancement of cognitiveabilities (e.g., attention or working memory) or affect per-ception directly. Conflicting evidence exists on whethertraditional antipsychotic medications improve cognitiveabilities (Allen et al. 1997; Sweeney et al. 1991) and theeffects on affect perception may be dose related (Corriganand Penn 1995).

How Does Attention Affect Emotion Perception? Theresults of this study showed that all four factors of atten-tion were significantly related to affect perception scores.A proposed theoretical model on the relationship betweenattention and affect perception, which suggests how atten-



tional problems might lead to impairments in affect per-ception, is presented below.

First, because affect perception is a dynamic processand subject to constant change, the ability to shift atten-tion from one facial expression to another is valuable.People impaired in this aspect of attention may becomestuck on one particular emotion, and miss subsequentshifts in emotion. This study showed that the Shift factoris most predictive of affect perception scores and isarguably the most important of the four factors. Second,the ability to sustain attention over time may be importantbecause people must constantly follow social dialogue todetect changes in emotional states. A person who con-stantly looks away and does not stay focused on the con-versation at hand is more likely to miss important socialand emotional cues. Bryson et al. (1997) argued that sus-tained attention leaves people ready to detect importantaspects of emotion. Third, encoding of the entire face (andpossibly other body cues) may also be important becausenarrowly focusing on a certain aspect of the face (e.g.,only the eyes or mouth) may lead to the wrong conclusionregarding the expressed emotion. The effect of encodingcan be seen in a study by Mandal and Palchoudhury(1989) in which persons with schizophrenia made moreerrors in affect perception when shown pictures of thewhole face (more complex encoding) than shown onlyparts of the face.

The influence of the Focus-Execute factor of atten-tion is more difficult to define. This factor is very simi-lar to processing speed. The important aspects of thisfactor are (1) the ability to identify an important feature(to focus) and (2) the ability to respond to it (to exe-cute). Those with delusions and schizophrenia tend toexamine more nonrelevant areas of the face (Phillipsand David 1997, 1998; Quirk 2000). Thus the personmay not be focusing on the most relevant areas of theface to obtain the most information. The ability of themotor system to generate emotionally appropriateresponses (through speech or nonverbal cues) is anotherkey component of this factor. Even when participants inthis study knew the correct emotion presented on thetape, that information may have been lost, or the emo-tion changed, if they did not respond quickly. Socially,if a person engaged in conversation does not respond(or is slow to respond) to an emotion, the other personmay lose interest, become upset, or end the interaction.A study by Mandal and Rai (1987) showed that peoplewith schizophrenia were slower to identify emotionalpictures than a group of people with anxiety disordersand a control group.

Each attentional factor arguably could impair affectperception. A summary of the proposed theoretical rela-tionship between attention and affect perception adopted

in this study can be found in figure 1. The fact that all fourfactors of attention were related to affect perception sug-gests that this skill can become impaired in many areas inpeople with schizophrenia.

Limitations. Sample size is a concern for the multipleregression analysis. Multiple regression proceduresshould have a participant-to-variable ratio of at least10:1 to ensure an appropriate degree of generalization(Diekoff 1992). For this study, the number of variablesin the regression (8) was relatively acceptable to thenumber of participants (n = 65). In addition, the combi-nation of the attentional measures (via PCA) into fourindependent factors further reduced the number of vari-ables in the multiple regression analysis and providedorthogonality among the attentional factors (Diekoff1992).

The sample has several limitations. First, the presentstudy did not include a control group of normal people.Recent findings by Kohler et al. (2000) showed that cog-nitive variables in normal people have little relationshipto affect perception and face-recognition performance,but a relationship between cognition and affect percep-tion did exist for people with schizophrenia. Becausecognition and affect perception may not be related innormal people, having a control group may be of littlebenefit to understanding the link between attention andaffect perception in schizophrenia, in which cognitivedeficits are more prominent (See Bryson et al. 1997 andKee et al. 1998 for similar designs). Second, the sampleused in this study could be placed at the severe end ofthe illness continuum based on length of current treat-ment, length of illness, and reading scores. The results ofthis study may differ from other studies with lessimpaired participants.

Some argue that affect perception tasks may bedesigned to highlight attentional problems in this popu-lation by having brief exposure times and reducing thenumber of visual and verbal cues available to the person(Bellack et al. 1996, 1999). In this study, the emotionalstimuli were presented for brief periods of time (10 to 15seconds), but the regression results of the BLERT (withcues/verbal material) and the FEIT (without cues) weresimilar. Therefore, even though stimulus presentationtimes were brief (which may be ecologically valid foremotional expression), there were no differences regard-ing the role of attention for the BLERT and the FEIT.

Finally, the measures of attention used in this studycannot be considered pure measures of attention; theymost likely involve other cognitive abilities as well. Infact, most psychological measures require a combinationof abilities for successful completion (Spreen andStrauss 1998). For example, the WCST has aspects of



Figure 1. Theoretical model of the relationship between attention and affect perception

Impact onAffect Perception

ShiftCannot Shift to Different

Emotions

Attention

SustainCannot Sustain

Attention to the Face

Focus-Execute

Slow to respond toemotions

Examine non-relevantareas of the face

Encode

Cannot take inall features of the

face

attention, problem solving, motor skills, and abstractreasoning (Spreen and Strauss 1998). Mirsky et al.(1991) stated that difference between shifting attentionand executive functioning (conceptual reasoning andflexibility) are unclear at present, but both have beenlinked to the frontal cortex. According to Posner andPetersen (1990), the Shift function is more related to eyemovements produced in the superior colliculus than tofrontal cortex functioning. Although the measures usedin this study have significant attentional components,they do have other aspects that should be considered inthe interpretation of these findings.

This study provided evidence for the role of atten-tional variables in affect perception. All four factors ofattention were significantly related to affect perceptionscores. More information is needed to clarify the complexrelationship between cognitive variables, symptoms, andmedication factors in affect perception. Each factor likelyaffects the others, and empirical research could helpuncover their individual and combined effects. This studyrepresents a first step in understanding how a single cog-nitive skill, attention, can affect social functioning. Wehope that future studies will further refine our understand-ing of attention and affect perception.



ReferencesAddington, J., and Addington, D. Facial affect recognitionand information processing in schizophrenia and bipolardisorder. Schizophrenia Research, 32:171-181, 1998.

Addington, J.; Addington, D.; and Gashore, L.Neurocognitive and social functioning in schizophreniaand other diagnoses. Schizophrenia Research,48:367-368,2001.

Allen, D.N.; Gilbertson, M.W.; van Kammen, D.R;Kelley, M.E.; Gurklis, J.A., Jr.; and Barry, E.J. Chronichaloperidol treatment does not affect structure of attentionin schizophrenia. Schizophrenia Research, 25:53-61,1997.

Archer, J.; Hay, D.C.; and Young, A.W. Face processingin psychiatric conditions. British Journal of ClinicalPsychology, 31:45-61, 1992.

Bell, M.; Bryson, G.; and Lysaker, P. Positive and nega-tive affect recognition in schizophrenia: A comparisonwith substance abuse and normal control subjects.Psychiatry Research, 73:73-82, 1997.

Bellack, A.S. Cognitive rehabilitation for schizophrenia:Is it possible? Is it necessary? Schizophrenia Bulletin,18(l):43-50, 1992.

Bellack, A.S.; Blanchard, J.J.; and Mueser, K.T. Cueavailability and affect perception in schizophrenia.Schizophrenia Bulletin, 22(3):535-544, 1996.

Bellack, A.S.; Gold, J.M.; and Buchanan, R.W. Cognitiverehabilitation for schizophrenia: Problems, prospects, andstrategies. Schizophrenia Bulletin, 25(2):257-274, 1999.

Benton, A.; Hamsher, K.; and Sivan, A.B. MultilingualAphasia Examination. 3rd ed. Iowa City, IA: AJAAssociates, 1994.

Benton, A.; Van Allen, M.; Hamsher, K.; and Levin, H.Test of Facial Recognition Manual. Iowa City, IA: BentonLaboratory of Neuropsychology, 1983.

Bezchlibnyk-Butler, K.Z., and Jeffries, J.J. ClinicalHandbook of Psychotropic Drugs. 12th ed. Seattle, WA:Hogrefe and Hubner, 2002.

Borod, J.C.; Martin, C.C.; Alpert, M.; Brozgold, A.; andWelkowitz, J. Perception of facial emotion in schizo-phrenic and right brain-damaged patients. Journal ofNervous and Mental Disease, 181:494-502, 1993.

Bruce, V., and Young, A. Understanding face recognition.British Journal of Psychology, 77:305-327, 1986.

Bryson, G.; Bell, M.; and Lysaker, P. Affect recognition inschizophrenia: A function of global impairment or a spe-cific cognitive deficit. Psychiatry Research, 71:105-113,1997.

Chapman, L.J., and Chapman, J.P. Problems in the mea-surement of cognitive deficit. Psychological Bulletin,79:380-385, 1973.

Chapman, L.J., and Chapman, J.P. The measurement ofdifferential deficit. Journal'of Psychiatric Research,14:303-311, 1978.

Chen, R.Y.L.; Chen, E.Y.H.; and Chan, C.K.Y. Verbal flu-ency in schizophrenia: Reduction in semantic store.Australian and New Zealand Journal of Psychiatry,34:43^8, 2000.

Cornblatt, B.A.; Lenzenweger, M.F.; Dworkin, R.H.; andErlenmeyer-Kimling, L. Childhood attentional dysfunctionspredict social deficits in unaffected adults at risk for schizo-phrenia. British Journal of Psychiatry, 161:59-64, 1992.

Corrigan, P.W., and Penn, D.L. The effects of antipsy-chotic and antiparkinsonian medication on psychosocialskill learning. Clinical Psychology: Science and Practice,2:251-262,1995.

Cramer, P.; Wegmann, M.; and O'Neill, M. Schizophreniaand the perception of emotions: How accurately do schiz-ophrenics judge the emotional states of others. BritishJournal of Psychiatry, 155:225-228, 1989.

Diekoff, G. Statistics for the Social and BehavioralSciences: Univariate, Bivariate, Multivariate. Dubuque,IA: Wm. C. Brown Publishers, 1992.

Dougherty, P.E.; Bartless, E.S.; and Izard, C.E. Responsesof schizophrenics to expressions of fundamental emo-tions. Journal of Clinical Psychology, 30:243-246, 1974.

Ekman, P. Pictures of facial affect. Palo Alto, CA:Consulting Psychologists Press, 1976.

First, M.B.; Spitzer, R.L.; Gibbon, M.; and Williams,J.B.W. Structured Clinical Interview for DSM-IV Axis IDisorders: Patient Edition. New York, NY: BiometricsResearch Department, 1995.

Freedman, L.R.; Rock, D.; Roberts, S.A.; Cornblatt, B.A.;and Erlenmeyer-Kimling, L. The New York High-RiskProject: Attention, anhedonia, and social outcome.Schizophrenia Research, 30:1-9, 1998.

Green, M.F. Cognitive remediation in schizophrenia: Is ittime yet? American Journal of Psychiatry, 150:178-187,1993.

Green, M.F. What are the functional consequences of neu-rocognitive deficits in schizophrenia? American Journalof Psychiatry, 153:321-330, 1996.

Heaton, R. The Wisconsin Card Sorting Test Manual.Odessa, FL: Psychological Assessment Resources, 1981.

Hogarty, G.E., and Flesher, S. Cognitive remediation inschizophrenia: Proceed...with caution! SchizophreniaBulletin, 18(l):51-57, 1992.



Ihnen, G.H.; Penn, D.L.; Corrigan, P.W.; and Martin, J.Social perception and social skill in schizophrenia.Psychiatry Research, 80:275-286, 1998.

Izard, C.E., The Face of Emotion. New York, NY:Appleton-Century-Crofts, 1971.

Kee, K.S.; Kern, R.S.; and Green, M.F. Perception ofemotion and neurocognitive functioning in schizophrenia:What's the link? Psychiatry Research, 81:57-65, 1998.

Kerr, S.L., and Neale, J.M. Emotion perception in schizo-phrenia: Specific deficit or further evidence of generalizedpoor performance? Journal of Abnormal Psychology,102:312-318, 1993.

Kline, J.S.; Smith, J.E.; and Ellis, H.C. Paranoid and non-paranoid schizophrenic processing of facially displayedaffect. Journal of Psychiatric Research, 26:169-182, 1992.

Kohler, C.G.; Bilker, W.; Hagendoorn, M.; Gur, R.E.; andGur, R.C. Emotion recognition deficit in schizophrenia:Association with symptomatology and cognition.Biological Psychiatry, 48:127-136, 2000.

Kremen, W.S.; Seidman, L.J.; Faraone, S.V.; Pepple, J.R.;and Tsuang, M.T. Attention/information processing fac-tors in psychotic disorders. Journal of Nervous andMental Disease, 180:89-93, 1992.

Lehman, A.H., and Steinwachs, D.M. TranslatingResearch Into Practice: The Schizophrenia PatientOutcomes Research Team (PORT) treatment recommen-dations. Schizophrenia Bulletin, 24(1): 1-10, 1998.

Lewis, S.F., and Garver, D.L. Treatment and diagnosticsubtype in facial affect recognition in schizophrenia.Journal of Psychiatry Research, 29:5-11, 1995.

Lukoff, D.; Nuechterlein, K.H.; and Ventura, J. Manualfor the Expanded Brief Psychiatric Rating Scale.Schizophrenia Bulletin, 12(4):594-602, 1986.

Magaro, P.A. The paranoid and the schizophrenic: A casefor distinct cognitive style. Schizophrenia Bulletin,7(4):632-661, 1981.

Mapou, R.L. A cognitive framework for neuropsychologi-cal assessment. In: Mapou, R.L., and Spector, J., eds.Clinical Neuropsychological Assessment: A CognitiveApproach. New York, NY: Plenum Press, 1995. pp.295-337.

Mandal, M.K., and Palchoudhury, S. Identifying the com-ponents of facial emotion and schizophrenia.Psychopathology, 22:295-300, 1989.

Mandal, M.K.; Pandey, R.; and Prassad, A.B. Facialexpressions of emotions and schizophrenia: A review.Schizophrenia Bulletin, 24(3):399-413, 1998.

Mandal, M.K., and Rai, A. Responses to facial emotionand psychopathology. Psychiatry Research, 20:317-323,1987.

Meltzer, H.Y., and McGurk, S.R. The effect of clozapine,risperidone, and olanzapine on cognitive function inschizophrenia. Schizophrenia Bulletin, 25(2):234-253,1999.

Mirsky, A.F.; Anthony, B.F.; Duncan, C.C.; Ahearn, M.B.;and Kellam, S.G. Analysis of the elements of attention: Aneuropsychological approach. Neuropsychology Review,2:109-145, 1991.

Mirsky, A.F.; Fantie, B.D.; and Tatman, J.E. Assessmentof attention across the lifespan. In: Mapou, R.L., andSpector, J., eds. Clinical Neuropsychological Assessment:A Cognitive Approach. New York, NY: Plenum Press,1995. pp. 17^18.

Morrison, R.L.; Bellack, A.S.; and Bashore, T.R.Perception of emotion among schizophrenic patients.Journal of Psychopathology and Behavioral Assessment,10:319-332, 1988a.

Morrison, R.L.; Bellack, A.S.; and Mueser, K.T. Deficitsin facial-affect recognition and schizophrenia.Schizophrenia Bulletin, 14(l):67-83, 1988fc.

Mueser, K.T.; Curran, P.J.; and McHugo, G.J. Factorstructure of the brief psychiatric rating scale in schizo-phrenia. Psychological Assessment, 9:196-204, 1997.

Mueser, K.T.; Doonan, R.; Penn, D.L.; Blanchard, J.J.;Bellack, A.S.; Nishith, P.; and DeLeon, J. Emotion recog-nition and social competence in chronic schizophrenia.Journal of Abnormal Psychology, 105:271-275, 1996.

Muzekari, L.H., and Bates, M.E. Judgment of emotionamong chronic schizophrenics. Journal of ClinicalPsychology, 32:662-666, 1977.

Nagamoto, H.T.; Adler, L.E.; Hea, R.A.; Griffith, J.M.;McRae, K.A.; and Freedman, R. Gating of auditory P50in schizophrenics: Unique effects of clozapine. BiologicalPsychiatry, 40:181-188, 1996.

Novic, J.; Daniel, L.; and Perline, R. Facial affect recogni-tion in schizophrenia: Is there a differential deficit?British Journal of Psychiatry, 39:774-777, 1984.

Nuechterlein, K.H. Vigilance in schizophrenia and relateddisorders. In: Steinhauer, S.R.; Gruzelier, J.H.; and Zubin,J., eds. Handbook of Schizophrenia: Neuropsychology,Psychophysiology, and Information Processing. Vol. 5.Amsterdam, The Netherlands: Elsevier Press, 1991. pp.397^33.

Penn, D.L., and Combs, D.R. Modification of affect per-ception deficits in schizophrenia. Schizophrenia Research,25:100-107,2000.

Penn, D.L.; Combs, D.R.; and Mohamed, S. Social cognitionand social functioning in schizophrenia. In: Corrigan, P.W.,and Penn, D.L., eds. Social Cognition and Schizophrenia.Washington, DC: APA Press, 2001. pp. 97-122.



Penn, D.L.; Combs, D.R.; Ritchie, M.; Francis, J.; Cassisi,J.; Morris, S.; and Townsend, M. Emotion recognition inschizophrenia: Further investigation of specific versusgeneralized deficit models. Journal of AbnormalPsychology, 109:512-516,2000.

Penn, D.L.; Corrigan, P.W.; Bentall, R.P.; Racenstein,J.M.; and Newman, L. Social cognition in schizophrenia.Psychological Bulletin, 121:114-132, 1997.

Penn, D.L., and Mueser, K.T. Research update on the psy-chosocial treatment of schizophrenia. American Journalof Psychiatry, 153:607-617, 1996.

Penn, D.L.; Spaulding, W.D.; Reed, D.; and Sullivan, M.The relationship of social cognition to ward behavior inchronic schizophrenia. Schizophrenia Research,20:327-335, 1996.

Phillips, M.L., and David, A.S. Visual scan paths areabnormal in deluded schizophrenics. Neuropsychologia,35:99-105, 1997.

Phillips, M.L., and David, A.S. Abnormal visual scanpaths: A psychophysiological marker of delusions inschizophrenia. Schizophrenia Research, 29:235-245,1998.

Posner, M.I., and Petersen, S.E. The attention system ofthe human brain. Annual Review of Neuroscience,13:25^2, 1990.

Quirk, S.W. Emotional response in schizophrenia:Relationships between disorganization symptoms, emo-tional report, and visual scanning. [Abstract]. DissertationAbstracts International, 60:4904, 2000.

Reitan, R.M., and Davidson, L.A. Clinical neuropsychol-ogy: Current status and applications. Washington, DC:V.H. Winston and Sons, 1974.

Salem, L.E.; Kring, A.M.; and Kerr, S.L. More evidencefor generalized poor performance in facial emotion per-ception in schizophrenia. Journal of AbnormalPsychology, 105:480-483, 1996.

Saykin, A.J.; Gur, R.C.; Gur, R.E.; Mozley, P.D.; Mozley,L.; Resnick, S.M.; Kester, B.; and Stafiniak, P.Neuropsychological function in schizophrenia: Selectiveimpairment in memory and learning. Archives of GeneralPsychiatry, 48:618-624, 1991.

Saykin, A.J., Shtasel, D.L., Gur, R.E.; Kester, D.B.;Mozley, L.H.; Stafiniak, P.; and Gur, R.C.Neuropsychological deficits in neuroleptic naive patientswith first-episode schizophrenia. Archives of GeneralPsychiatry, 51:124-131, 1994.

Schneider, F.; Gur, R.C; Gur, R.E.; and Shtasel, D.L.Emotional processing in schizophrenia: Neurobehavioralprobes in relation to psychopathology. SchizophreniaResearch, 17:67-75, 1995.

Schwartz, B.L.; Rosse, R.B.; and Deutsch, S.L. Toward aneuropsychology of memory in schizophrenia.Psychopharmacology Bulletin, 28:341-351, 1992.

Sohlberg, M.M., and Mateer, C.A. Introduction toCognitive Rehabilitation: Theory and Practice. NewYork, NY: Guilford Press, 1989.

Spreen, A., and Strauss, E. A Compendium ofNeuropsychological Tests: Administration, Norms, andCommentary. 2nd ed. New York, NY: Oxford UniversityPress, 1998.

Steinhauer, S.R.; Zubin, J.; Condray, R.; Shaw, D.B.;Peters, J.L.; and Van Kammen, D.P. Electrophysiologicaland behavioral signs of attentional disturbance in schizo-phrenics and their siblings. In: Tamminga, C.A., andSchultz, S.C., eds. Advances in Neuropsychiatry andPsychopharmacology, Volume 1: Schizophrenia Research.New York, NY: Raven Press, 1991.

Stip, E., and Lussier, I. The effect of risperidone on cogni-tion in patients with schizophrenia. Canadian Journal ofPsychiatry, 41:35^0, 1996.

Strauss, M.E. Relations of symptoms to cognitive deficitsin schizophrenia. Schizophrenia Bulletin, 19(2):215-231,1993.

Sweeney, J.A.; Keilp, J.G.; Haas, G.L.; Hill, J.; andWeiden, P. Relationships between medication treatmentsand neuropsychological test performance in schizophre-nia. Psychiatry Research, 37:297-308, 1991.

Thompson, B. Stepwise regression and stepwise discrimi-nant anaylsis need not apply here: A guidelines editorial.Educational and Psychological Measurement, 55:525-534, 1995.

Toomey, R.; Schuldberg, D.; Corrigan, P.W.; and Green,M.F. Nonverbal social perception and symptomatology inschizophrenia. Schizophrenia Research, 53: 83-91, 2002.

Wilkinson, G.S. Wide Range Achievement TestAdministration Manual. 3rd ed. Wilmington, DE: WideRange, 1993.

Zalewski, C ; Johnson-Selfridge, M.T.; Ohriner, S.;Zarrella, K.; and Seltzer, J.C. A review of neuropsycho-logical differences between paranoid and nonparanoidschizophrenia patients. Schizophrenia Bulletin,24(1): 127-145, 1998.


the role of attention in affect perception: an examination of mir

Documents