Raphael Gillett

How to do a meta-analysis.

Meta-analysis is a statistical tool for estimating the mean and variance of underlying population effects from a collection of empirical studies addressing ostensibly the same research question. Meta-analysis has become an increasing popular and valuable tool in psychological research, and major review articles typically employ these methods. This article describes the process of conducting meta-analysis: selecting articles, developing inclusion criteria, calculating effect sizes, conducting the actual analysis (including information on how to do the analysis on popular computer packages such as IBM SPSS and R) and estimating the effects of publication bias. Guidance is also given on how to write up a meta-analysis.

Personality and coping: A context for examining celebrity worship and mental health

The adaptational-continuum model of personality and coping suggests a useful context for research areas that emphasize both personality and coping. The present paper used Fergusons (2001) model integrating personality and coping factors to further conceptualize findings around celebrity worship. Three hundred and seventy-two respondents completed measures of celebrity worship, personality, coping style, general health, stress, positive and negative affect and life satisfaction. Celebrity worship for intense-personal reasons is associated with poorer mental heath and this relationship can be understood within the dimensions of neuroticism and a coping style that suggests disengagement. Such findings suggest the utility of examining the relationship between celebrity worship and mental health within both personality and coping variables, which have practical implications for understanding and addressing mental health problems that may occur as the result of engaging in celebrity worship for intense-personal reasons.

Mindfulness, functioning and catastrophizing after multidisciplinary pain management for chronic low back pain.

Summary Relationships between self‐reports of mindfulness, catastrophizing, disability and depression were explored before and after a multidisciplinary cognitive‐behavioural intervention for chronic pain. Abstract We examined mindfulness in people with chronic low back pain who were attending a multidisciplinary pain management programme. Participants completed questionnaires at baseline (n = 116) and after a 3‐month cognitive‐behaviourally informed multidisciplinary intervention (n = 87). Self‐reported mindfulness was measured before and after the intervention, and relationships were explored between mindfulness, disability, affect and pain catastrophizing. Mindfulness increased following participation in the intervention, and greater mindfulness was predictive of lower levels of disability, anxiety, depression and catastrophizing, even when pain severity was controlled. Mediator analyses suggested that the relationship between mindfulness and disability was mediated by catastrophizing. It is possible that cognitive‐behavioural interventions and processes can affect both catastrophizing and mindfulness.

Perceptions and prevalence of stalking in a male sample

Abstract Most research into stalking has focused on males as perpetrators and females as victims. The present study investigated firstly how males perceive the crime of stalking and secondly the prevalence of stalking and other intrusive activities in a male population. A sample of 210 British men was asked to indicate which of a continuum of 42 behaviours they believed to represent stalking. Clear subtypes of perceived stalking were identified, and these did not significantly differ from those produced by a female sample in earlier work (Sheridan, Davies and Boon, 2001). Thus, male and females hold similar views on which activities do and do not constitute ‘stalking’, Respondents were then asked to indicate whether they had personal experience of the same 42 behaviours and if they had, to provide free narrative on their ‘worst experience’. As expected, males reported substantially less experience of intrusive behaviour than females, and just 5.2% were judged to have suffered ‘stalking’. Whilst this figure is significantly below estimates made for females, it still indicates a notable risk for males and suggests that complaints of stalking made by males should be taken as seriously as those made by females.

‘Stalking’ – Seeking the victim's perspective

Abstract The law has experienced great difficulty in defining ‘stalking’ and framing legislation to deal with the problem. There has been little research as to how the public in general and potential victims in particular view this phenomenon. In an effort to explore this issue, 80 women aged 18 to 55 years were invited to classify which of a range of intrusive behaviours were exemplars of stalking. A cluster analysis of the classification data showed a clear distinction between those which participants felt to be stalking and non-stalking related acts. Further, a number of distinct sub-categories of perceived stalking behaviours were suggested by the analysis. Participants were also asked to indicate whether they had first hood experience of any of these behaviours and to describe their worst personal experience of ‘stalking’. Analysis of this data suggested that instances of what our respondents perceived as ‘stalking’ was widespread in the sample and that some of its more serious manifestations would, if reported, have led to criminal charges. The difficulties of framing legislation, however, to define all the acts which our respondents saw as ‘stalking’ remain.

Understanding the Factors That Affect the Severity of Juvenile Stranger Sex Offenses The Effect of Victim Characteristics and Number of Suspects

Understanding factors that affect the severity of a juvenile-stranger sexual assault has implications for crime prevention, and potentially, the assessment and treatment of juvenile sex offenders. This study investigated how victim characteristics and the number of suspects affected the use of physical violence and weapons and the occurrence of penetration in 495 allegations of sexual assault committed by juveniles against strangers. Statistically significant interactions between victim age and gender were found for occurrence of penetration and use of violence. Differences in offense characteristics were also found between offenses with varying victim-suspect age differences. When comparing the rate of penetration in the presence and absence of violence, little change was observed for lone suspects. However, the rate of penetration increased significantly for groups in the presence of physical violence, suggesting that violence in this context may be more expressive than instrumental. Theoretical explanations and practical implications are considered.

An average power criterion for sample size estimation

The use of an empirical estimate of effect size in the standard sample size formula introduces variability in the power to detect the true underlying effect. A useful criterion for judging the performance of a sample size estimation procedure, when there is uncertainty about effect size, is that the average power supplied to those studies in a discipline seeking a particular level of power should reach the nominal level. The common informal practice of substituting an effect size estimate from an earlier study in the traditional sample size formula fails to meet this criterion, yielding an average power that falls below the nominal level, often by a considerable margin. An alternative method of sample size determination that satisfies the average power criterion is outlined.

Post hoc power analysis

The objective of a post hoc review of statistical power is to determine the average power of tests used in a particular field. Because most theories in psychology are not sufficiently advanced to provide an exact prediction of effect size, post hoc reviews invariably calculate the power of a test in terms of the tests ability to detect the expected effect size as estimated from the data. However, to achieve their stated objective, such reviews should instead compute the expected power of a test over the range of likely effect sizes. That is, current practice commits the statistical error of confusing a function of an expectation with the expectation of a function. In consequence, post hoc power reviews overestimate the average power of statistical tests, often by a substantial margin

Celebrity worship, addiction and criminality

Abstract Two studies assessed the relationship between celebrity worship and (i) addiction (n=1359) and (ii) criminality (n=2158). Overall Celebrity Attitude Scale (CAS) scores correlated positively with Eysenck Personality Questionnaire – Revised (EPQ-R) Addiction and Criminality sub-scale scores. In further support of the absorption–addiction model of celebrity worship (McCutcheon, Lange, & Houran, 2002, Conceptualization and measurement of celebrity worship. British Journal of Psychology, 93, 67–87.), addiction correlated positively with one component of celebrity worship, and criminality correlated positively with all four components produced by the current work. Different types of celebrity worship were associated with preferences for celebrities from particular domains. For instance, those scoring highly on the “Deleterious Imitation” component favoured music celebrities, as opposed to political figures. It was concluded that pathological celebrity worshippers are seeking a personal identity and are drawn to particular celebrities. Imitating these celebrities can have negative consequences for the worshipper.

Pitfalls in assessing research performance by grant income

The strategy of judging the quality of scientific research by the level of funding it attracts is critically examined. It is argued that an indes such as per captita research income, which is based on grant-giver peer review, yields an unsatisfactory measure of scientific performance. It fails to fulfil a basic requirement of a performance indicator, namely, that it should relate outputs to inputs. It has intrinsically low validity, and is strongly confounced with a variety of extraneous factors that are unrelated to research performance.