COGNITIVE NEUROPSYCHIATRY 2011, 16 (5), 461
472

Metacognition and depressive realism: Evidence for the level-of-depression account Nicholas C. Soderstrom1, Deana B. Davalos1, and Susana M. Va´zquez2 1

Department of Psychology, Colorado State University, Fort Collins, CO, USA 2 Ponce School of Medicine, Puerto Rico

Introduction. The present study examined the relationship between metacognition (i.e., ‘‘thinking about thinking’’) and depression. More specifically, the depressive realism hypothesis (Alloy & Abramson, 1979), which posits that depressed people have a more accurate view of reality than nondepressed people, was tested. Methods. Nondepressed, mildly depressed, and moderately depressed individuals predicted their memory performance by making judgements of learning after each studied item. These predictions were then compared with actual performance on a free recall task to assess calibration, an index of metacognitive accuracy. Results and conclusions. Consistent with the depressive realism hypothesis, mild depression was associated with better calibration than nondepression. However, this ‘‘sadder but wiser’’ phenomenon appears to only exist to point, as moderate depression and nondepression showed no calibration differences. Thus, the level-ofdepression account of depressive realism is supported.

Keywords: Depression; Depressive realism; Judgements of learning; Metacognition; Metamemory.

Correspondence should be addressed to Nicholas C. Soderstrom, Department of Psychology, Colorado State University, Fort Collins, CO 80523-1876 USA. E-mail: nicholas.soderstrom@ colostate.edu This research was sponsored by the National Science Foundation (NSF) Research Experiences for Undergraduates (REU) and Department of Defense Awards to Stimulate and Support Undergraduate Research Experiences (ASSURE) Programs, NSF Grant SES-0552876 to Edward L. DeLosh. The authors would like to thank Alexa Pizer for her contributions to this project. # 2011 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business http://www.psypress.com/cogneuropsychiatry DOI: 10.1080/13546805.2011.557921

462

´ ZQUEZ SODERSTROM, DAVALOS, VA

INTRODUCTION Depressive realism is the notion that depressed individuals have a more realistic, or less distorted, view of reality as compared to their nondepressed counterparts. Using contingency-judgement tasks, Alloy and Abramson (1979) were the first to provide empirical evidence for this ‘‘sadder but wiser’’ phenomenon. In their study, individuals either pressed or did not press a button before witnessing a light turning on; after multiple trials, participants reported their perceived control over this outcome. Across a variety of actual contingency relationships between the behaviour (i.e., pressing the button) and the outcome (i.e., light turning on), depressed students consistently reported more accurate control judgements than nondepressed students. Since the time of this seminal work, a substantial amount of research has been dedicated to the study of depressive realism, sparking debates as to whether this phenomenon actually exists (for reviews, see Allan, Siegal, & Hannah, 2007; Dobson & Franche, 1989). Controversy surrounds this phenomenon, in large part because it runs counter to traditional cognitive conceptions of depression that emphasise the role of negative perceptions in this disorder (e.g., Beck, 1967). From a depressive realism perspective, nondepressed individuals are those with unrealistic or biased perceptions; those with depression have a firmer grasp of reality. This viewpoint has interesting clinical implications, including the notion that realistic thinking should be questioned as a mental health criterion. As stated by Dobson and Franche, ‘‘if persons with low self-esteem and depression are actually those well anchored in reality, and the rest of the population is to some extent distorting reality in a positive direction in order to maintain a state of non-depression, then one of the criteria for mental health may have to be the maintenance of positively distorted perceptions of self and of the world’’ (1989, p. 420). Thus, this approach would suggest that treatment of depression should focus on, among other things, the promotion of positively biased perceptions. Three major, nonmutually exclusive explanations have been put forward to account for depressive realism (see Ackerman & DeRubies, 1991, for a review). First, the schematic processing account suggests that, whereas nondepressed individuals process information in a more positive light, thus leading to positive illusions of control, depressed individuals process information in a more negative/neutral way, leading to more ‘‘realistic’’ judgements (Dykman, Abramson, Alloy, & Hartlage, 1989). That is, individuals are thought to possess certain self-schemata by which they define themselves and process information (Beck, 1967). Thus, the lack of positive illusions in depressed individuals as a result of their negative selfschemata leads to depressive realism. Second, the motivational bias account posits that positively biased judgements in nondepressed individuals reflect

METACOGNITION AND DEPRESSIVE REALISM

463

their motivation to maintain high self-esteem, whereas those with depression are thought to have significantly reduced motivation of this sort, resulting in less biased judgements (Bradley, 1978). This account is concerned with the maintenance of self-esteem, not the motivation to improve it; thus, it follows that nondepressed individuals would show more positively biased judgements than those with depression. Last, the level-of-depression account argues that depressive realism reflects more realistic thinking in mildly depressed or dysphoric individuals, not those with moderate or severe depression (Reuhlman, West, & Pasahow, 1985). Indeed, it may be the case that moderate and severe depression, like nondepression, is associated with major distortions in thinking, whereas mild depression may be related to a reduction or absence of such distortions. Linking this account with the schemata processing account described earlier, mildly depressed individuals’ self-schemata might contain a balance of both positive and negative material, leading to relatively less biased judgements. To further explore the notion of depressive realism, the current study extended this concept into the realm of metacognition. Metacognition refers to ‘‘thinking about thinking’’, and captures one’s ability to both monitor his/ her learning (i.e., self-reflect) and control his/her behaviour as a function of this monitoring (for a review, see Koriat, 2007; see also Wells & Matthews, 1996). To illustrate this relationship, Nelson and Narens (1990) distinguished between the object level and the meta level. The meta level oversees objectlevel processes, modifying them when appropriate. For example, a student studying for an upcoming exam may come across a concept that is difficult to understand. While reading and encoding the information (i.e., object-level processes), the student may become aware that the material is not well understood (i.e., metalevel process). Consequently, the student may choose to remedy the situation by rereading the material. This feedback loop between the object level and the metalevel is critical in more serious situations in which one needs to be acutely aware of his or her own strengths and weaknesses in comprehension (see Dunning, Heath, & Suls, 2004). Although the role of metacognitive beliefs in depression has been articulated from a clinical perspective (Corcoran & Segal, 2008), and low cognitive confidence has been linked to this disorder (see Wells, 2008), empirical evidence speaking to the relationship between metacognition and depression is lacking. For example, in one of the few studies on this topic, Cipolli, Neri, de Vreese, Pinelli, Rubichi, and Lalla (1996) found a negative correlation between depression and metacognitive abilities in an older adult population. However, in this case metacognitive abilities were scored on the basis of survey questions, not on experimental procedures. Furthermore, this study did not include younger adults. Because other studies on metacognition and depression have similar limitations (e.g., Ponds & Jolles, 1996), we

464

´ ZQUEZ SODERSTROM, DAVALOS, VA

investigated this relationship in younger adults, whereby metacognitive abilities were measured experimentally. The particular experimental methodology used in the current study is related to one subcategory of metacognition called metamemory, which refers to the knowledge one possesses regarding his or her own memory processes and capabilities. The standard method of experimentally investigating metamemory, and the method used currently, is to elicit judgements of learning (JOLs; see Nelson, 1996). Here, participants are asked during a study phase to assess the likelihood of remembering a particular item on a later test. That is, people are asked to predict their future memory performance. JOLs are typically made on a 0
100% scale and are compared to actual memory performance to determine metacognitive accuracy. It is worth noting here that in the present study, metacognitive accuracy is denoted by the correspondence between mean JOL and mean memory performance, a measure called calibration (Lichtenstein, Fischhoff, & Phillips, 1982), which allows for an overall assessment of under- and overconfidence. Employing the abovementioned experimental procedure and operationalising its outcome measure of calibration (i.e., under/overconfidence) as an index of how ‘‘realistic’’ one is regarding his/her memory abilities, we were able to probe the issue of depressive realism with an alternative approach. Further, by including multiple depression levels, we were able to get a more nuanced look at the depression
metamemory relationship, thereby allowing us to test the level-of-depression account of depressive realism discussed earlier. Recall that this account states that mild depression, compared to nondepression and moderate
severe depression, is associated with the most realistic judgements (Reuhlman et al., 1985). Thus, pertaining to the current study, the hypothesis derived from this account predicts that, compared to both nondepression and moderate depression, mild depression will be associated with the best calibration (i.e., the most ‘‘realistic’’ memory predictions).

METHOD Participants Participants were undergraduate students from a mid-sized university. Given that students were recruited from freshman- and sophomore-level psychology courses, rather than selected from a clinical sample, data collection was difficult and spanned three semesters. Students were recruited for this experiment using a mass screening method. All volunteers were required to complete the Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977) before being invited to participate. Of the 347 students who were prescreened, 97 (65 female, 32 male) were selected on the basis of their

METACOGNITION AND DEPRESSIVE REALISM

465

CES-D scores. Ages ranged from 18 to 34 years, with a mean age of 20.5 years. Participants were recruited on campus (in-class or via fliers) and received either course credit (n 61) or $10.00 compensation (n 36) for their participation. All participants were treated in accordance with the ‘‘Ethical Principles of Psychologists and Code of Conduct’’ (American Psychological Association, 2002) and the study had the approval of the Institutional Review Board at Colorado State University.

Depression classification The CES-D (Radloff, 1977) was administered as our measure of depression. The CES-D is one of the most frequently used and well-validated measures of depression. This short, self-report scale includes 20 items in which the person indicates how often certain thoughts or behaviours (e.g., ‘‘I was bothered by things that usually don’t bother me’’) were experienced during the last week. Each answer is given on a 5-point scale ranging from ‘‘rare or none of the time’’ to ‘‘most or all of the time’’. Based on the scores obtained from the CES-D, participants were classified as belonging to one of three groups: nondepression, mild depression, and moderate depression (see Table 1 for demographic information and CES-D scores broken down by depression group). This scale has been used to assess depressive symptomatology on both a continuum and to identify categories of depression (Santor, Zuroff, Ramsay, Cervantes, & Palacios, 1995). Numerous studies, which include a wide variety of participants (e.g. varying in age, medical condition, and psychological comorbid conditions), have supported the cutoff score of 24 to determine moderate
severe depression (Roberts, Lewinsohn, & Seeley, 1991; van Voorhees et al., 2008). However, the traditional cutoff score of 16 to indicate mild depression has been questioned in adolescent and college-aged students. Specifically, research has supported the use of higher cutoff scores and/or the use of a median split (Roberts et al., 1991; Tercyak & Audrain, 2002). Given this, we first split our sample on the median (19), giving us two groups that corresponded to what we considered TABLE 1 Mean age, number of females/males, and average CES-D scores as a function of depression group Depression group Nondepression Mild depression Moderate depression

Age

Females/males

CES-D

20.81 (3.25) 20.05 (2.50) 20.16 (1.80)

36/22 13/7 16/3

13.29 (4.74) 21.60 (1.39) 29.32 (3.11)

Standard deviations are reported in parentheses where appropriate.

466

´ ZQUEZ SODERSTROM, DAVALOS, VA

nondepressed and depressed. Next, we further split the depressed group on the score of 24 to distinguish between mild depression and moderate depression. Thus, we operationally defined nondepression as a score of 19 or below (n 58), mild depression as a score of 20
23 (n20), and moderate depression as a score of 24 and above (n 19).1 Most of our analyses were conducted using these classifications; however, we also included a supplemental analysis treating depression as a continuous variable to avoid the controversies surrounding the current depression cutoff points (see Results).

Materials Stimuli included 36 affectively valenced, single words (12 positive, 12 neutral, 12 negative). The positive and negative words were selected from those used in previous research (e.g., Denny & Hunt, 1992), whereas the neutral words were selected from the Affective Norms for English Words (ANEW) stimulus collection (Bradley & Lang, 1999). The length of the words and their frequency of occurrence were equated. Two randomised, fixed study lists were created for purposes of counterbalancing. Stimuli were presented on computers with SuperLab software.

Procedure Participants were tested individually. After providing informed consent, a short demographic questionnaire and the CES-D were administered. Upon completion of these questionnaires, the experimenter read the instructions for the memory experiment. Participants were informed that they would be studying words for an upcoming test, and that they would be predicting their memory performance for each word. Each word was studied for 3 s, after which time the JOL prompt was presented (‘‘Chances of Recall? 0%
100%’’), replacing the studied word. JOLs were made aloud on a scale from 0% (‘‘surely will not recall’’) to 100% (‘‘surely will recall’’) and were self-paced. Participants were encouraged to use the entire range of the scale when giving their predictions. Once each JOL was made, the experimenter recorded the response and pressed a key to advance the screen to the next word
JOL sequence. The study phase was completed once all 36 words had been given JOLs. After the study phase, a short distractor task (a few minutes in duration) was performed to displace the studied items from short-term 1

We originally categorised the individuals scoring above 24 on the CES-D as ‘‘severely’’ depressed. Following the suggestion of a reviewer, we dropped the label of severe and replaced it with moderate. Indeed, this label seems more appropriate, especially given that this is not a clinical sample.

METACOGNITION AND DEPRESSIVE REALISM

467

memory. This consisted of a trail making test in which letters and numbers were connected in ascending order. Finally, a free recall task was performed in which participants wrote down, on a lined piece of paper, as many words as possible from the study list. For counterbalancing purposes, half of the participants completed the CES-D before the memory experiment, and the other half completed it after the memory experiment.

RESULTS Metacognitive accuracy Our primary interest concerned level of depression as it relates to calibration, an index of metacognitive accuracy that quantifies the correspondence between mean prediction and mean performance. Thus, for each depression group we subtracted mean performance (converted to a percentage) from mean JOL, which yielded group calibration bias scores. These data are shown in Figure 1. A one-way analysis of variance (ANOVA) was then conducted to determine if calibration bias differed as a function of level of depression. A significant effect was found, F(2, 94) 6.87, MSE 274.02, p B.05, h2p .13. Post hoc comparisons using Tukey’s HSD test revealed that mild depression (M11.24, SE 3.72) was associated with lower calibration bias than both nondepression (M22.87, SE 2.18, Cohen’s d 1.18) and moderate depression (M30.60, SE 3.72, Cohen’s d 1.00). No

Figure 1. Mean calibration bias as a function of level of depression (error bars represent standard error). Note that zero represents perfect calibration.

468

´ ZQUEZ SODERSTROM, DAVALOS, VA

difference was found when comparing the nondepression group to the moderate depression group. Taken together, these data reveal that, although all groups were overconfident (i.e., mean JOLs were higher than percentage correct recall), mild depression was associated with the least amount of overconfidence, indicating that this group was the best calibrated. To further investigate this finding, we conducted two subsequent analyses. First, we conducted an analysis of covariance (ANCOVA), adding the age and sex of participants as covariates to control for these variables. The reliable difference of calibration bias associated with depression still held, F(2, 92) 7.25, MSE 243.98, p B.01, h2p .24; thus, differences in the age and sex of our participants cannot explain these differences.2 We also conducted an analysis treating depression as a continuous variable given that controversy often surrounds the use of cutoff scores when classifying depression. Consistent with our previous findings, this analysis yielded a significant quadratic term, F(2, 94) 3.01, p B.05, R2 .06 (y31.56
1.59x.05x2). Thus, these supplemental analyses provide converging evidence for our conclusions regarding the relationship between depression and metacognitive accuracy.

Free recall To investigate in what ways level of depression is related to overall memory performance alone, we conducted a one-way ANOVA with free recall as the dependent variable, which revealed a significant effect, F(2, 94) 3.80, MSE 8.73, pB.05, h2p .08. Tukey’s HSD post hoc comparisons revealed that the mean number of recalled words in the moderate depression group (M 8.05, SD2.12) was significantly lower than the amount recalled by the mild depression group (M 10.65, SD3.36, Cohen’s d 0.93). No differences were found when comparing these groups to the nondepression group (M 9.21, SD3.04). Not surprisingly, these results indicate that moderate depression is associated with worse memory performance than mild depression.

JOLs To rule out a possible anchoring explanation*that all groups anchor their JOLs on a similar point on the judgement scale, whereas performance 2

Studies investigating the effects of depression on various outcome measures often control for level of mood during the experimental procedure as well. Given that the depression scores for present analyses were derived from the CES-D administration during testing, the assignment of individuals to different levels of depression does reflect their mood during the experiment.

METACOGNITION AND DEPRESSIVE REALISM

469

differs*we conducted a one-way ANOVA with mean JOL as the dependent measure. This analysis yielded a significant difference, F(2, 94) 3.52, MSE 259.72, pB.05, h2p .07. Tukey’s HSD post hoc comparisons revealed that the mean JOL for the mild depression group (M40.82, SD15.79) was lower than that of the moderate depression group (M 54.43, SD15.87, Cohen’s d 0.86). No differences were found when comparing these groups to the nondepression group (M48.45, SD16.30). Thus, it is not the case that all groups anchored their JOLs in similar ways.

DISCUSSION The primary purpose of the current study was to investigate the depressive realism hypothesis*that depressed individuals have a more accurate view of reality than nondepressed individuals*as it relates to metacognitive accuracy. To do so, we categorised individuals as nondepressed, mildly depressed, or moderately depressed and had them perform a metamemory task in which item-by-item memory predictions (i.e., JOLs) were elicited and compared with actual performance. This prediction
performance comparison, known as calibration, is an index of metacognitive accuracy that allows for an assessment of how ‘‘realistic’’ one is in regards to predicting memory performance. Our results indicate that mild depression is associated with better calibration than both nondepression and moderate depression. That is, mildly depressed individuals showed a closer correspondence between their average JOL and memory performance as compared to their nondepressed and moderately depressed counterparts, suggesting that mild depression is related to more ‘‘realistic’’ thinking in this domain. Moreover, because calibration did not differ between nondepression and moderate depression, the levels-of-depression account of depressive realism is supported. This account posits that mild depression is associated with more realistic thinking, whereas both nondepression and moderate
severe depression are characterised by greater thought distortions (Ruehlman et al., 1985). Indeed, we show that, although all depression groups were overconfident, mild depression was associated with significantly less overconfidence than the other groups. Other theories of depressive realism, such as the schematic processing account (Dykman et al., 1989) and the motivational bias account (Bradley, 1978), only distinguish between nondepression and depression when explaining this phenomenon (see Introduction); thus, although our results are not totally inconsistent with these theories (we found that mild depression was related to more realistic thinking than nondepression), they match more closely with those predictions made by the levels-of-depression

470

´ ZQUEZ SODERSTROM, DAVALOS, VA

account. Specifically, relative to nondepression and moderate depression, mild depression showed more accurate (i.e., more ‘‘realistic’’) memory judgements. If mild depression is indeed characterised by a more accurate view of reality, a natural question arises: Why? It has been argued elsewhere (e.g., Alloy & Abramson, 1988) that nondepression may be characterised by a tendency to process information in an overly positive manner in order to maintain high levels of self-esteem. From this perspective, then, our results reflect an adaptive process at work among our nondepression group. It should be noted, however, that, compared to our nondepression group, our moderate depression group was equally overconfident in their memory abilities. Surely, maintaining high self-esteem among those with moderate depression is not responsible for their overconfidence, as a defining characteristic of this condition is the presence of low self-esteem. Thus, it follows that the overconfidence shown in this group may be explained differently. For example, high levels of depression have been associated with faulty information processing in general, which, in turn, may lead to thought distortions (Austin, Mitchell, & Goodwin, 2001). Mild depression, then, perhaps represents a middle ground between nondepression and moderate depression insofar as it may be associated with an absence of both high selfesteem maintenance and faulty information processing.3 Of course, these ideas are speculative, in need of empirical confirmation. In addition to exploring the relationship between metacognition and depression, we also examined how memory performance alone was related to depression. Previous research has suggested that, for the most part, depression has a detrimental effect on memory performance (for reviews, see Blaney, 1986; Burt, Zembar, & Niederehe, 1995). Consistent with this research, we found that those individuals with moderate depression recalled fewer words overall compared to those with mild depression, a finding that perhaps helps to validate our depression categorisation. Although we feel that the experimental paradigm used in the current study provides a fresh, promising approach to the study of depressive realism, we recognise that there are limitations specific to our study. For example, because we categorised depression on the basis of scores from a depression questionnaire, it is possible that this phenomenon may not generalise to other types of depression, particularly clinical depression (but see Ghaemi, Sachs, Baldassano, & Truman, 1997, for a clinical study consistent with our results). Furthermore, there are notable limitations 3

We do not want to leave the reader with the impression that this is the only possible explanation for depressive realism. Indeed, recent work suggests that this phenomenon might be explained in terms of contextual processing differences (Msetfi, Murphy, & Simpson, 2007) and/or passive responses (Blanco, Matute, & Vadillo, 2009) among those with depression.

METACOGNITION AND DEPRESSIVE REALISM

471

associated with the use of self-reports such as the CES-D, including their subjective nature. Finally, the fact that our participants were undergraduates and mostly female speaks to the need to extend this research to other populations with varying educational backgrounds. We believe that this line of research represents a fruitful area of future study. The notion of depressive realism is now over 30 years old, dating back to Alloy and Abramson’s (1979) now-classic contingency-judgement experiments. Since that time, reconciling this phenomenon with major theories of depression (e.g., Beck, 1967) that link this condition with thought distortions has proven difficult. We have provided evidence, however, that seems consistent with both of these views. That is, we have shown that mild depression is associated with more realistic thinking than both nondepression and moderate depression (consistent with depressive realism), and that moderate depression is related to a greater level of thought distortion than mild depression (consistent with traditional depression theories). Thus, perhaps the adage ‘‘sadder but wiser’’ should be changed to ‘‘mildly sadder but wiser’’. Manuscript received 17 June 2010 Revised manuscript received 10 November 2010 First published online 9 March 2011

REFERENCES Ackerman, R., & DeRubeis, R. J. (1991). Is depressive realism real? Clinical Psychology Review, 11, 565
584. Allan, L. G., Siegel, S., & Hannah, S. (2007). The sad truth about depressive realism. Quarterly Journal of Experimental Psychology, 60, 482
495. Alloy, L. B., & Abramson, L. Y. (1979). Judgment of contingency in depressed and nondepressed students: Sadder but wiser? Journal of Experimental Psychology: General, 108, 441
485. Alloy, L. B., & Abramson, L. Y. (1988). Depressive realism: Four theoretical perspectives. In L. B. Alloy (Ed.), Cognitive processes in depression (pp. 223
265). New York, NY: Guilford Press. American Psychological Association. (2002). Ethical principles of psychologists and code of conduct. The American Psychologist, 57, 1060
1073. Austin, M. P., Mitchell, P., & Goodwin, G. M. (2001). Cognitive deficits in depression: Possible implications for functional neuropathology. British Journal of Psychiatry, 178, 200
206. Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects. New York, NY: Harper & Row. Blanco, F., Matute, H., & Vadillo, M. A. (2009). Depressive realism: Wiser or quieter? Psychological Record, 59, 551
562. Blaney, P. H. (1986). Affect and memory: A review. Psychological Bulletin, 99, 229
246. Bradley, G. W. (1978). Self-serving biases in the attribution process: A re-examination of the fact or fiction question. Journal of Personality and Social Psychology, 36, 56
71. Bradley, M. M., & Lang, P. J. (1999). Affective norms for English words (ANEW): Stimuli, instruction manual and affective ratings (Tech. Rep. No. C-1). Gainesville, FL: The Center for Research in Psychophysiology, University of Florida. Burt, D. B., Zembar, M. J., & Niederehe, G. (1995). Depression and memory impairment: A metaanalysis of the association, its pattern, and specificity. Psychological Bulletin, 117, 285
305.

472

´ ZQUEZ SODERSTROM, DAVALOS, VA

Cipolli, C., Neri, M., de Vreese, L. P., Pinell, P., Rubichi, S., & Lalla, M. (1996). The influence of depression on memory and metamemory in the elderly. Archives of Gerontology and Geriatrics, 23, 111
127. Corcoran, K. M., & Segal, Z. V. (2008). Metacognition in depressive and anxiety disorders: Current directions. International Journal of Cognitive Therapy, 1, 33
44. Denny, E. B., & Hunt, R. R. (1992). Affective valence and memory in depression: Dissociation of recall and fragment completion. Journal of Abnormal Psychology, 101, 575
580. Dobson, K., & Franche, R.-L. (1989). A conceptual and empirical review of the depressive realism hypothesis. Canadian Journal of Behavioural Science, 21, 419
433. Dunning, D., Heath, C., & Suls, J. M. (2004). Flawed self-assessment: Implications for health, education, and the workplace. Psychological Science in the Public Interest, 5, 69
106. Dykman, B. M., Abramson, L. Y., Alloy, L. B., & Hartlage, S. (1989). Processing ambiguous and unambiguous feedback by depressed and non-depressed college students: Schematic biases and their implications for depressive realism. Journal of Personality and Social Psychology, 56, 431
435. Ghaemi, S. N., Sachs, G. S., Baldassano, C. F., & Truman, C. J. (1997). Insight in seasonal affective disorder. Comprehensive Psychiatry, 38, 345
348. Koriat, A. (2007). Metacognition and consciousness. In P. Zelazo (Ed.), The Cambridge handbook of consciousness (pp. 289
325). New York, NY: Cambridge University Press. Lichtenstein, S., Fischhoff, B., & Phillips, L. D. (1982). Calibration of probabilities: The state of the art to 1980. In D. Kahneman, P. Slovic, & A. Tversky (Eds.), Judgment under uncertainty: Heuristics and biases (pp. 306
344). New York, NY: Cambridge University Press. Msetfi, R. M., Murphy, R. A., & Simpson, J. (2007). Depressive realism and the effect of intertrial interval on judgments of zero, positive, and negative contingencies. Quarterly Journal of Experimental Psychology, 60, 461
481. Nelson, T. O. (1996). Consciousness and metacognition. The American Psychologist, 51, 102
116. Nelson, T. O., & Narens, L. (1990). Metamemory: A theoretical framework and some new findings. In G. H. Bower (Ed.), The psychology of learning and motivation (pp. 125
173). New York, NY: Academic Press. Ponds, R. W. H. M., & Jolles, J. (1996). Memory complaints in elderly people: The role of memory abilities, metamemory, depression, and personality. Educational Gerontology, 22, 341
357. Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Assessment, 1, 305
401. Roberts, R. E., Lewinsohn, P. M., & Seeley, J. R. (1991). Screening for adolescent depression: A comparison of depression scales. Journal of American Academy of Child and Adolescent Psychiatry, 30, 58
66. Ruehlman, L. S., West, S. G., & Pasahow, R. J. (1985). Depression and evaluative schemata. Journal of Personality, 53, 46
92. Santor, D. A., Zuroff, D., Ramsay, J. O., Cervantes, P., & Palacios, J. (1995). Examining scale discriminability in the BDI and CES-D as a function of depressive severity. Psychological Assessment, 7, 131
139. Tercyak, K. P., & Audrain, J. (2002). Psychosocial correlates of alternate tobacco product use during early adolescence. Preventative Medicine, 35, 193
198. Van Voorhees, B. W., Paunesku, D., Gollan, J., Kuwabara, S., Reinecke, M., & Basu, A. (2008). Predicting future risk of depressive episode in adolescents: The Chicago adolescent depression risk assessment (CADRA). Annals of Family Medicine, 6, 503
511. Wells, A. (2008). Metacognitive therapy: Cognition applied to regulating cognition. Behavioural and Cognitive Psychotherapy, 36, 651
658. Wells, A., & Matthews, G. (1996). Modelling cognition in emotional disorder: The S-REF model. Behaviour Research and Therapy, 34, 881
888.

Copyright of Cognitive Neuropsychiatry is the property of Psychology Press (UK) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.