This article was downloaded by: [Mount Royal University] On: 01 May 2015, At: 06:07 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Cognitive Neuropsychiatry Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/pcnp20
The effect of paranoia on the judging of harmful events a
a
a
a
Daniel Freeman , Nicole Evans , Emma Černis , Rachel Lister & b
Graham Dunn a
Department of Psychiatry, University of Oxford, Oxford, UK
b
Centre for Biostatistics, Institute of Population Health, University of Manchester, Manchester, UK Published online: 27 Oct 2014.
Click for updates To cite this article: Daniel Freeman, Nicole Evans, Emma Černis, Rachel Lister & Graham Dunn (2015) The effect of paranoia on the judging of harmful events, Cognitive Neuropsychiatry, 20:2, 122-127, DOI: 10.1080/13546805.2014.976307 To link to this article: http://dx.doi.org/10.1080/13546805.2014.976307
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Downloaded by [Mount Royal University] at 06:07 01 May 2015
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
Cognitive Neuropsychiatry, 2015 Vol. 20, No. 2, 122–127, http://dx.doi.org/10.1080/13546805.2014.976307
The effect of paranoia on the judging of harmful events Daniel Freemana*, Nicole Evansa, Emma Černisa, Rachel Listera and Graham Dunnb a
Department of Psychiatry, University of Oxford, Oxford, UK; bCentre for Biostatistics, Institute of Population Health, University of Manchester, Manchester, UK
Downloaded by [Mount Royal University] at 06:07 01 May 2015
(Received 27 March 2014; accepted 7 October 2014) Objective. Social psychological research has indicated that intentional harm may be perceived as causing greater damage than unintentional harm. It has been proposed that this harm magnification is a consequence of a need to blame, condemn and punish (“blame motivation”). The objective of the current study was to replicate these findings and to test whether such judgements about harmful events are associated with the level of an individual’s paranoia. Method. Three hundred adults read a scenario in which a head of a company causes a reduction in employees’ pay. Participants were randomly allocated to versions in which the outcome of the executive’s action was intended or unintended. Ratings were made of intent, harm caused and blame motivation. Participants also completed assessments of paranoia and anxiety. Results. Intentional harm was judged as causing worse outcomes than unintentional harm, explaining a small amount of variance in harm scores. Paranoia moderated judgements of intent and blame motivation but not the degree of harm caused; high paranoia, relative to low paranoia, was associated with the unintentional scenario generating higher attributions of intent and blame and the intentional scenario generating lower attributions of intent and blame. Anxiety levels did not affect judgements. Conclusions. The study supports the theory that there is a reasoning bias that magnifies the consequences of intentional harm relative to unintentional harm. In the initial judgement about intent, people with paranoia are less accurate in their use of contextual information. Keywords: Paranoia; delusions; attributions
When harm occurs, judgements swiftly follow. In a recent series of innovative experimental studies, Ames and Fiske (2013) show that the consequences of identical harm are judged differently depending upon the perception of the perpetrator’s intent. Intended harm is judged as causing greater damage than unintended, but identical, harm. The desire to blame, condemn and punish (“blame motivation”) accounted for this inflation of intended harm. The researchers’ view was that blame motivation is more easily satisfied if the effects of harm are emphasised. This social psychological work on harm highlights a number of complex judgements – concerning others’ intentions, attribution of blame and the actual harm caused – potentially subject to bias. We expected the accuracy of such judgements to be affected by an individual’s level of paranoia. Paranoia concerns erroneous judgements that harm is going to occur to the self and that the perpetrator has the intention to cause it (Freeman & Garety, 2000). Paranoia is *Corresponding author. Email: [email protected] © 2014 Taylor & Francis
Downloaded by [Mount Royal University] at 06:07 01 May 2015
Cognitive Neuropsychiatry
123
specifically about harm to the self, but we considered it plausible that there would be a generalised bias to see harm as intentional and that there would be a stronger reaction towards it. Ames and Fiske (2013) tested their hypothesis about estimates of harm in a series of online studies in the USA. In their first study, 80 people read a scenario in which the head of a small company loses the employees’ pay because of a bad investment. The participants were randomised to an intentional or unintentional version of the vignette. In the intentional harm condition, the participants read that the loss was deliberate in order to get the employees working harder. In the unintentional harm condition, the participants were simply told that the loss was due to chance. The intentional harm condition led to greater ratings of actual harm. The increase in the estimates of harm was statistically accounted for by an increase in blame motivation. In the current study, we sought not only to replicate the study by Ames and Fiske (2013) but also to develop the understanding of the findings by identifying individual differences that may influence the effects. We were interested in whether the judgements are affected by levels of paranoia. We predicted that higher levels of paranoia would be associated with higher ratings of perpetrator intent, that the tendency to blame would be greater and that estimates of actual harm caused would therefore be higher. The first prediction is consistent with evidence that paranoia is associated with externalising attributions for negative events occurring to the self (see review by Garety & Freeman, 2013). The two further predictions are untested.
Method Participants Three hundred adults (aged 18 or over) took part. Participants were recruited from the general population using local radio adverts, leafleting of local areas and posters. There were 190 females and 110 men, with a mean age of 45.7 (SD = 17.3). The ethnicity of the sample was predominately White (n = 281). The marital status were married/civil partnership (n = 116), single (n = 92), cohabiting (n = 49), divorced (n = 28) and widowed (n = 15). Highest levels of education achieved were none/GCSE (n = 28), A level (n = 64), degree (n = 111) and postgraduate degree (n = 97). The employment status of the group were full-time employed (n = 114), retired (n = 63), part-time employed (n = 42), students (n = 33), self-employed (n = 34) or unemployed (n = 14). Assessments Green et al. Paranoid Thoughts Scale (GPTS) The GPTS is a 32-item measure of paranoid thinking over the past month. Part A assesses ideas of reference and Part B assesses ideas of persecution (Green et al., 2008). Each item is rated on a 5-point scale. Higher scores indicate greater levels of paranoid thinking. The internal consistency of the scale in the current study was very high (Cronbach’s alpha = .97). Depression Anxiety Stress Scales: Anxiety (DASS Anxiety) The DASS Anxiety subscale was used, comprising 14 items with a 0–3 scale (0 = did not apply to me at all, 3 = applied to me very much; Lovibond & Lovibond, 1995). Higher
124
D. Freeman et al.
scores indicate higher levels of current anxiety. The scale has been shown to be reliable and valid in a large UK non-clinical population (Crawford & Henry, 2003). Judgement task
Downloaded by [Mount Royal University] at 06:07 01 May 2015
The study was completed online using Qualtrics software, with randomisation to condition carried out automatically by the programme. The vignette used was that developed by Ames and Fiske (2013). The two versions of the vignette describe identical harms caused to employees by the head of a company, Terrance Smith. The vignettes only differ in the descriptions of his intentions. In the intentional condition, participants read that: Terrance decided to invest in a software development project because he was confident that it would fail. He thought his employees might become lazy now that the company was starting to do well, and that they would work harder to increase profits in the future if they made less money now. The investment failed.
In the unintentional condition, participants read that: “Terrance decided to invest in a software development project because it seemed like a good opportunity for the company. Unfortunately, not all investments are successful. The investment failed”. After reading the vignette, participants answered five questions, using sliding scales from 0 to 100: 1. How much harm would you say Terrance’s (the CEO’s) investment in the software development project did to the employees? 2. To what extent do you think Terrance deserves blame for making the investment? 3. To what extent do you think Terrance should be morally condemned for making the investment? 4. To what extent do you want Terrance to be punished for making the investment? 5. To what extent did Terrance intend for his employees to lose money? As in the original report, Questions 2–4 were combined to form a blame motivation subscale. The internal reliability of this subscale was good (intentional condition Cronbach’s alpha = .84, unintentional condition Cronbach’s alpha = .86). Analysis All analyses were carried out using SPSS (IBM, 2011). To test for replication of the original Ames and Fiske (2013) study, we compared the ratings in the two conditions (intentional harm, unintentional harm) for estimates of intent, harm caused and blame motivation. We used analysis of covariance, controlling for socio-demographic variables (gender, age, ethnicity, education, marital status, employment status). To test for moderation of scores by levels of paranoia, we constructed separate univariate models for each rating, entering condition (intentional, unintentional) and paranoia score as independent variables and testing for an interaction. If there was no significant interaction then we tested for a main effect of paranoia. In these analyses, we controlled for the socio-demographic variables. The GPTS scores were transformed (log to the base 10) to reduce significant skew. We also checked whether judgements were moderated by levels of anxiety to assess specificity to paranoia. All significance testing were two-tailed. There were no missing data.
Cognitive Neuropsychiatry
125
Results Replication
Downloaded by [Mount Royal University] at 06:07 01 May 2015
The ratings from each condition are shown in Table 1. The two scenarios were clearly perceived as intentional and unintentional. The intentional condition led to greater estimates of the harm caused, but this explained only a small amount of the variance. The intentional scenario was also clearly associated with greater blame motivation scores. Ratings for intent, harm caused and blame motivation were not associated with level of education (p > .1). Moderation by paranoia The attribution of intent across condition was moderated by level of paranoia (p = .001). The estimate of harm caused was not (p = .161). Total blame motivation was moderated by levels of paranoia (p = .006), and this was explained particularly by interactions with blame (p = .018) and condemnation (p = .001) but not punishment (p = .098). To interpret the interactions, the total sample was divided into low- and high paranoia based upon a median split. It can be seen that high paranoia, relative to low paranoia, was associated with seeing less intent in the intentional scenario but greater intent in the unintentional scenario (Table 2). Consistent with this, high paranoia, relative to low paranoia, is associated with less blame motivation in the intentional condition but greater blame motivation in the unintentional condition. Despite this, there are indications in the mean scores displayed in Table 2 that harm caused was rated as slightly higher in both conditions in the high paranoia group, but the main effect of paranoia in an analysis of covariance (removing the interaction term) was not statistically significant (p = .098). For Table 1. Mean scores for the scenario ratings. Intentional group (n = 155) Mean (SD) Intention Harm caused Blame Condemn Punish Blame motivation total
64.3 51.7 86.1 71.3 54.1 211.6
Unintentional group (n = 145) Mean (SD)
(36.6) (26.1) (23.5) (29.4) (34.1) (76.4)
5.2 46.6 57.1 35.5 30.0 122.6
df
(11.3) (26.1) (33.6) (34.3) (32.5) (90.0)
16, 16, 16, 16, 16, 16,
283 283 283 283 283 283
F
p-value
Partial η2
298.1 4.3 74.4 88.6 38.4 81.5
Cognitive Neuropsychiatry Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/pcnp20
The effect of paranoia on the judging of harmful events a
a
a
a
Daniel Freeman , Nicole Evans , Emma Černis , Rachel Lister & b
Graham Dunn a
Department of Psychiatry, University of Oxford, Oxford, UK
b
Centre for Biostatistics, Institute of Population Health, University of Manchester, Manchester, UK Published online: 27 Oct 2014.
Click for updates To cite this article: Daniel Freeman, Nicole Evans, Emma Černis, Rachel Lister & Graham Dunn (2015) The effect of paranoia on the judging of harmful events, Cognitive Neuropsychiatry, 20:2, 122-127, DOI: 10.1080/13546805.2014.976307 To link to this article: http://dx.doi.org/10.1080/13546805.2014.976307
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Downloaded by [Mount Royal University] at 06:07 01 May 2015
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
Cognitive Neuropsychiatry, 2015 Vol. 20, No. 2, 122–127, http://dx.doi.org/10.1080/13546805.2014.976307
The effect of paranoia on the judging of harmful events Daniel Freemana*, Nicole Evansa, Emma Černisa, Rachel Listera and Graham Dunnb a
Department of Psychiatry, University of Oxford, Oxford, UK; bCentre for Biostatistics, Institute of Population Health, University of Manchester, Manchester, UK
Downloaded by [Mount Royal University] at 06:07 01 May 2015
(Received 27 March 2014; accepted 7 October 2014) Objective. Social psychological research has indicated that intentional harm may be perceived as causing greater damage than unintentional harm. It has been proposed that this harm magnification is a consequence of a need to blame, condemn and punish (“blame motivation”). The objective of the current study was to replicate these findings and to test whether such judgements about harmful events are associated with the level of an individual’s paranoia. Method. Three hundred adults read a scenario in which a head of a company causes a reduction in employees’ pay. Participants were randomly allocated to versions in which the outcome of the executive’s action was intended or unintended. Ratings were made of intent, harm caused and blame motivation. Participants also completed assessments of paranoia and anxiety. Results. Intentional harm was judged as causing worse outcomes than unintentional harm, explaining a small amount of variance in harm scores. Paranoia moderated judgements of intent and blame motivation but not the degree of harm caused; high paranoia, relative to low paranoia, was associated with the unintentional scenario generating higher attributions of intent and blame and the intentional scenario generating lower attributions of intent and blame. Anxiety levels did not affect judgements. Conclusions. The study supports the theory that there is a reasoning bias that magnifies the consequences of intentional harm relative to unintentional harm. In the initial judgement about intent, people with paranoia are less accurate in their use of contextual information. Keywords: Paranoia; delusions; attributions
When harm occurs, judgements swiftly follow. In a recent series of innovative experimental studies, Ames and Fiske (2013) show that the consequences of identical harm are judged differently depending upon the perception of the perpetrator’s intent. Intended harm is judged as causing greater damage than unintended, but identical, harm. The desire to blame, condemn and punish (“blame motivation”) accounted for this inflation of intended harm. The researchers’ view was that blame motivation is more easily satisfied if the effects of harm are emphasised. This social psychological work on harm highlights a number of complex judgements – concerning others’ intentions, attribution of blame and the actual harm caused – potentially subject to bias. We expected the accuracy of such judgements to be affected by an individual’s level of paranoia. Paranoia concerns erroneous judgements that harm is going to occur to the self and that the perpetrator has the intention to cause it (Freeman & Garety, 2000). Paranoia is *Corresponding author. Email: [email protected] © 2014 Taylor & Francis
Downloaded by [Mount Royal University] at 06:07 01 May 2015
Cognitive Neuropsychiatry
123
specifically about harm to the self, but we considered it plausible that there would be a generalised bias to see harm as intentional and that there would be a stronger reaction towards it. Ames and Fiske (2013) tested their hypothesis about estimates of harm in a series of online studies in the USA. In their first study, 80 people read a scenario in which the head of a small company loses the employees’ pay because of a bad investment. The participants were randomised to an intentional or unintentional version of the vignette. In the intentional harm condition, the participants read that the loss was deliberate in order to get the employees working harder. In the unintentional harm condition, the participants were simply told that the loss was due to chance. The intentional harm condition led to greater ratings of actual harm. The increase in the estimates of harm was statistically accounted for by an increase in blame motivation. In the current study, we sought not only to replicate the study by Ames and Fiske (2013) but also to develop the understanding of the findings by identifying individual differences that may influence the effects. We were interested in whether the judgements are affected by levels of paranoia. We predicted that higher levels of paranoia would be associated with higher ratings of perpetrator intent, that the tendency to blame would be greater and that estimates of actual harm caused would therefore be higher. The first prediction is consistent with evidence that paranoia is associated with externalising attributions for negative events occurring to the self (see review by Garety & Freeman, 2013). The two further predictions are untested.
Method Participants Three hundred adults (aged 18 or over) took part. Participants were recruited from the general population using local radio adverts, leafleting of local areas and posters. There were 190 females and 110 men, with a mean age of 45.7 (SD = 17.3). The ethnicity of the sample was predominately White (n = 281). The marital status were married/civil partnership (n = 116), single (n = 92), cohabiting (n = 49), divorced (n = 28) and widowed (n = 15). Highest levels of education achieved were none/GCSE (n = 28), A level (n = 64), degree (n = 111) and postgraduate degree (n = 97). The employment status of the group were full-time employed (n = 114), retired (n = 63), part-time employed (n = 42), students (n = 33), self-employed (n = 34) or unemployed (n = 14). Assessments Green et al. Paranoid Thoughts Scale (GPTS) The GPTS is a 32-item measure of paranoid thinking over the past month. Part A assesses ideas of reference and Part B assesses ideas of persecution (Green et al., 2008). Each item is rated on a 5-point scale. Higher scores indicate greater levels of paranoid thinking. The internal consistency of the scale in the current study was very high (Cronbach’s alpha = .97). Depression Anxiety Stress Scales: Anxiety (DASS Anxiety) The DASS Anxiety subscale was used, comprising 14 items with a 0–3 scale (0 = did not apply to me at all, 3 = applied to me very much; Lovibond & Lovibond, 1995). Higher
124
D. Freeman et al.
scores indicate higher levels of current anxiety. The scale has been shown to be reliable and valid in a large UK non-clinical population (Crawford & Henry, 2003). Judgement task
Downloaded by [Mount Royal University] at 06:07 01 May 2015
The study was completed online using Qualtrics software, with randomisation to condition carried out automatically by the programme. The vignette used was that developed by Ames and Fiske (2013). The two versions of the vignette describe identical harms caused to employees by the head of a company, Terrance Smith. The vignettes only differ in the descriptions of his intentions. In the intentional condition, participants read that: Terrance decided to invest in a software development project because he was confident that it would fail. He thought his employees might become lazy now that the company was starting to do well, and that they would work harder to increase profits in the future if they made less money now. The investment failed.
In the unintentional condition, participants read that: “Terrance decided to invest in a software development project because it seemed like a good opportunity for the company. Unfortunately, not all investments are successful. The investment failed”. After reading the vignette, participants answered five questions, using sliding scales from 0 to 100: 1. How much harm would you say Terrance’s (the CEO’s) investment in the software development project did to the employees? 2. To what extent do you think Terrance deserves blame for making the investment? 3. To what extent do you think Terrance should be morally condemned for making the investment? 4. To what extent do you want Terrance to be punished for making the investment? 5. To what extent did Terrance intend for his employees to lose money? As in the original report, Questions 2–4 were combined to form a blame motivation subscale. The internal reliability of this subscale was good (intentional condition Cronbach’s alpha = .84, unintentional condition Cronbach’s alpha = .86). Analysis All analyses were carried out using SPSS (IBM, 2011). To test for replication of the original Ames and Fiske (2013) study, we compared the ratings in the two conditions (intentional harm, unintentional harm) for estimates of intent, harm caused and blame motivation. We used analysis of covariance, controlling for socio-demographic variables (gender, age, ethnicity, education, marital status, employment status). To test for moderation of scores by levels of paranoia, we constructed separate univariate models for each rating, entering condition (intentional, unintentional) and paranoia score as independent variables and testing for an interaction. If there was no significant interaction then we tested for a main effect of paranoia. In these analyses, we controlled for the socio-demographic variables. The GPTS scores were transformed (log to the base 10) to reduce significant skew. We also checked whether judgements were moderated by levels of anxiety to assess specificity to paranoia. All significance testing were two-tailed. There were no missing data.
Cognitive Neuropsychiatry
125
Results Replication
Downloaded by [Mount Royal University] at 06:07 01 May 2015
The ratings from each condition are shown in Table 1. The two scenarios were clearly perceived as intentional and unintentional. The intentional condition led to greater estimates of the harm caused, but this explained only a small amount of the variance. The intentional scenario was also clearly associated with greater blame motivation scores. Ratings for intent, harm caused and blame motivation were not associated with level of education (p > .1). Moderation by paranoia The attribution of intent across condition was moderated by level of paranoia (p = .001). The estimate of harm caused was not (p = .161). Total blame motivation was moderated by levels of paranoia (p = .006), and this was explained particularly by interactions with blame (p = .018) and condemnation (p = .001) but not punishment (p = .098). To interpret the interactions, the total sample was divided into low- and high paranoia based upon a median split. It can be seen that high paranoia, relative to low paranoia, was associated with seeing less intent in the intentional scenario but greater intent in the unintentional scenario (Table 2). Consistent with this, high paranoia, relative to low paranoia, is associated with less blame motivation in the intentional condition but greater blame motivation in the unintentional condition. Despite this, there are indications in the mean scores displayed in Table 2 that harm caused was rated as slightly higher in both conditions in the high paranoia group, but the main effect of paranoia in an analysis of covariance (removing the interaction term) was not statistically significant (p = .098). For Table 1. Mean scores for the scenario ratings. Intentional group (n = 155) Mean (SD) Intention Harm caused Blame Condemn Punish Blame motivation total
64.3 51.7 86.1 71.3 54.1 211.6
Unintentional group (n = 145) Mean (SD)
(36.6) (26.1) (23.5) (29.4) (34.1) (76.4)
5.2 46.6 57.1 35.5 30.0 122.6
df
(11.3) (26.1) (33.6) (34.3) (32.5) (90.0)
16, 16, 16, 16, 16, 16,
283 283 283 283 283 283
F
p-value
Partial η2
298.1 4.3 74.4 88.6 38.4 81.5
