Do health beliefs predict health behaviour? A follow-up analysis of breast self-examination.

In a previous paper, we suggested that women who attended classes on breast self-examination showed significant improvements in behaviour, and that part of the reason was a change in their beliefs. Further analysis of our data confirms the original conclusion, and indicates that the perceived value of doing the behaviour is a more important predictor than perceived vulnerability to cancer.

Do health beliefs predict health behaviour? An analysis of breast self-examination.

The purpose of this paper is to explore the relationships between health beliefs and health behaviour. The study we report was conducted as part of a national campaign to evaluate the benefits of breast self-examination in the early detection of breast cancer, and the purpose of our analysis was to test the Health Belief Model of Becker and his colleagues. Three groups of women were investigated--278 who accepted an invitation to attend self-examination classes and were taught the techniques in detail, 262 who declined the invitation and 594 controls to whom no classes were offered--and beliefs and self-reported behaviour were measured shortly before the classes took place and again a year later. The campaign, it emerged, produced marked changes in both beliefs and behaviour, but the relationships between beliefs and behaviour were much weaker than the model had led us to expect and accounted for no more than a quarter of the variance. Alternative models are considered, notably Fishbein and Ajzen's Theory of Reasoned Action, and the paper concludes with some suggestions for improving health campaigns.

Inequalities in pregnancy outcome: a review of psychosocial and behavioural mediators.

The purpose of this paper is to review the literature on psychosocial factors in pregnancy outcome and to present a model which attempts to integrate the findings theoretically. There are four sections. The first presents published data on the incidence of early childhood mortality and low birth weight. Changes over time and differences between countries are noted and attention is drawn to the marked inequalities between occupational groups in the British data. The second section reviews the evidence that a variety of psychosocial risk factors influence pregnancy outcome, notably social, emotional, cognitive and behavioural factors. The third section develops the theme of inequalities and examines theories which have been advanced to account for the differences in adult mortality. We argue that material deprivation goes some way towards explaining inequalities in pregnancy outcome, but that any proper account will have to explain the links between inputs and outcomes--the processes and mechanisms by which material deprivation is translated into observable mortality and morbidity. In the concluding section, we argue that some of the principal links are the psychosocial risk factors described in the second section, and we present a model which traces the pathways of mediation.

Motorcycling safety research: a review of the social and behavioral literature.

The literature on motorcycling safety research is reviewed, and it is argued that there have been two main periods. The first, spanning the 1970's, was based on accident analysis, and the main objective was to identify and control factors that contribute to the severity of motorcycling accidents. The main concerns were to reduce head and brain injuries through safety helmets, to reduce multi-vehicle collisions through daytime use of headlamps, and to reduce drink-riding. The second period, the 1980's shifted the emphasis of research to what might be called 'riding analysis'--that is, analysis of the process of motorcycle riding. Particular attention was paid to skills testing, training evaluation and perceived risk. Now, in the early 1990's, a third period is developing, in which the rider is seen as 'active agent'. The theoretical basis of the new research has come from the models of social psychology, and the main concern is to use riders' beliefs and attitudes about safe riding to predict their behaviour on the roads and so their accident involvement. The three periods of research are reviewed in turn, and the paper concludes with an outline of the key research issues that remain to be addressed.

Attendance at cancer screening in the wake of widespread adverse publicity surrounding test results.

The largest error ever discovered in cervical smear test results was reported by an East Kent Hospitals' NHS Trust in 1996. To test whether the incident would have an impact on a separate NHS screening programme within the affected area, 1000 women who were due to be called for x ray mammography were asked how confident they would be about the accuracy of their mammogram result. Most women reported that they would be confident, but significantly fewer unconfident than confident women subsequently kept their appointment for breast screening. Attendance and non-attendance at breast screening in the area did not differ from previous screening rounds, suggesting that the error in the cervical cytology results had not adversely affected a separate screening programme.

Age and experience in motorcycling safety.

Official casualty statistics show that young motorcyclists are more likely than older motorcyclists to be killed or seriously injured on the roads. We address two main issues: might the statistics be attributable to inexperience rather than youth; and might accidents be associated with particular patterns of behaviour which may themselves be predictable from riders' beliefs? From a national prospective survey of over 4000 riders in the U.K., the data showed that youth played a much greater role than inexperience, and that accidents were associated with a particular pattern of behaviour, namely a willingness to break the law and violate the rules of safe riding, which was predictable from the riders' beliefs measured 12 months earlier. The implications of the findings are discussed both for theory and for policy and practice, and particular emphasis is placed on suggestions for modifying training courses.

Belief importance and the theory of planned behaviour: comparing modal and ranked modal beliefs in predicting attendance at breast screening.

OBJECTIVE: The objective of this study was, to compare the predictive utility of two measures of the attitude, subjective norm and perceived behavioural control constructs of the theory of planned behaviour (TPB; Ajzen, 1991) in predicting intention and subsequent attendance at breast screening. One construct was based on a modal set of underlying beliefs; the other was based on the three beliefs from each construct considered by the participant to be the most important. METHOD: We used a prospective, longitudinal design using a postal questionnaire at Time 1 and objective attendance data from screening records at Time 2. Questionnaires were sent to 1657 women from southeast England due to be invited for X-ray mammography under the UK's National Health Service Breast Screening Programme. After evaluating a set of modal behavioural, normative and control beliefs, women were asked to select the three beliefs they saw as the most important for them, and to rank them. The products of these three beliefs formed the 'important' (vs. the modal) measures. RESULTS: The 'important' attitude and subjective norm measures showed similar associations with direct measures and were equivalent to the modal measures in predicting intention and attendance at screening. The modal control construct was marginally more robust than the 'important' one in its association with the direct measure of control, and with intention. Key individual important beliefs that predicted intention and behaviour were identified. CONCLUSIONS: Measuring belief importance can help more fully identify the structures underlying attitude, subjective norm and perceived behavioural control, and can provide useful information when the TPB is used as the basis for intervention to help change behaviour.

Perceptions of risk in motorcyclists: unrealistic optimism, relative realism and predictions of behaviour.

In the first phase of a prospective investigation, a national sample of motorcyclists completed a postal questionnaire about their perceptions of risk, their behaviour on the roads and their history of accidents and spills. In the second phase a year later, they reported on their accident history and behaviour over the preceding 12 months. A total of 723 respondents completed both questionnaires. Four sets of findings are reported. First, the group as a whole showed unrealistic optimism: on average, respondents believed themselves to be less at risk than other motorcyclists of an accident needing hospital treatment in the next year. Second, optimism was tempered by 'relative realism', in that respondents who were young and inexperienced saw themselves as more at risk than other motorcyclists, as did riders who reported risky behaviours on the road. Third, there was some evidence of debiasing by personal history, in that having a friend or a relative who had been killed or injured on the roads was associated with perceptions of absolute risk of injury or death--though there were no effects on comparative risk and no effects on any of the judgments of a history of accidents of one's own. Finally, there was good evidence that perceptions of risk predicted subsequent behaviour, though generally in the direction not of precaution adoption but of precaution abandonment: the greater the perceived risk at time 1, the more frequent the risky behaviour at time 2. The implications of the findings are discussed, and possible interpretations are suggested.

Emotion and adherence to treatment in people with asthma: an application of the emotional stroop paradigm.

This study explored whether an emotional Stroop paradigm might represent an appropriate means of assessing individuals' emotional representations of asthma. In addition, the opportunity was taken to investigate whether emotional representations of asthma, as assessed by this method, were associated with adherence to inhaled preventative medication. An asthma Stroop task was devised which comprised three sets of stimuli: asthma symptom words, general negative words, and neutral words. Three groups of participants were compared on their performance on this task: individuals with asthma, individuals without asthma, and individuals without asthma who had been primed about the condition. It was found that individuals with asthma experienced significantly more interference when colour-naming the asthma symptom words, but not when colour-naming the general negative words. Furthermore, their performance on the asthma Stroop task was associated with self-reported adherence levels. Specifically, individuals who reported the highest and lowest levels of adherence displayed more interference when colour-naming the asthma symptom words than individuals with intermediate levels of adherence. It is concluded that the emotional Stroop paradigm might provide an objective and sensitive means of assessing individuals' emotional representations of illness. Additionally, it is proposed that emotional responses to illness should be assessed and included in research designed to explain health behaviours and, furthermore, that such research should not assume that any relationship between emotional representations and health behaviours will be linear.

Visual communication and the content and style of conversation.

Previous research suggests that visual communication plays a number of important roles in social interaction. In particular, it appears to influence the content of what people say in discussions, the style of their speech, and the outcomes they reach. However, the findings are based exclusively on comparisons between face-to-face conversations and audio conversations, in which subjects sit in separate rooms and speak over a microphone-headphone intercom which precludes visual communication. Interpretation is difficult, because visual communication is confounded with physical presence, which itself makes available certain cues denied to audio subjects. The purpose of this paper is to report two experiments in which the variables were separated and content and style were re-examined. The first made use of blind subjects, and again compared the face-to-face and audio conditions. The second returned to sighted subjects, and examined four experimental conditions: face-to-face; audio; a curtain condition in which subjects sat in the same room but without visual communication; and a video condition in which they sat in separate rooms and communicated over a television link. Neither visual communication nor physical presence proved to be critical variable. Instead, the two sources of cues combined, such that content and style were influenced by the aggregate of available cues. The more cueless the settings, the more task-oriented, depersonalized and unspontaneous the conversation. The findings also suggested that the primary effect of cuelessness is to influence verbal content, and that its influence on both style and outcome occurs indirectly, through the mediation of content.

Discomfort and pain during mammography: description, prediction, and prevention.

OBJECTIVE: To identify the nature of pain and discomfort experienced during mammography and how it can be ameliorated. DESIGN: Questionnaire survey before invitation for mammography and immediately after mammography. Responses before screening were related to experience of discomfort. SETTING: Health district in South East Thames region. SUBJECTS: 1160 women aged 50-64 invited routinely for screening; 774 completed first questionnaire, of whom 617 had mammography. 597 completed the second questionnaire. MAIN OUTCOME MEASURES: Reported discomfort and pain, comparisons of discomfort with that experienced during other medical procedures, qualitative description of pain with adjective checklist. RESULTS: 35% (206/597) of the women reported discomfort and 6% (37/595) pain. 10 minutes after mammography these figures were 4% (24/595) and 0.7% (4/595) respectively. More than two thirds of women ranked having a tooth drilled, having a smear test, and giving blood as more uncomfortable than mammography. The most important predictor of discomfort was previous expectation of pain (discomfort was reported by 21/32 (66%) women who expected pain and 186/531 (35%) who did not). Discomfort had little effect on satisfaction or intention to reattend. CONCLUSIONS: The low levels of reported pain and discomfort shortly after mammography and the favourable comparisons with other investigations suggest that current procedures are acceptable. Since two thirds of the women experienced less pain than expected health education and promotion must ensure that accurate information is made available and publicized.

An implementation intentions intervention to increase uptake of mammography.

BACKGROUND AND PURPOSE: This study tested an implementation intentions intervention to increase uptake in the United Kingdom's National Health Service Breast Screening Programme. The intervention asked women to plan how they would overcome up to 3 previously identified barriers to attending. METHODS: In a randomized controlled trial, 2,082 participants were allocated to an intervention condition, an assessment-only condition, or a nonassessment control condition. The intervention condition was designed to help women plan how to change their appointment, how to arrange transport, and how to negotiate time off work. The assessment-only condition controlled for the possibility that completing a questionnaire about mammography might in itself influence attendance, and the nonassessment condition was a control against any effect on attendance that mere contact with the research team might have. RESULTS: Mean age of respondents was 56.1 years, and 99.4% were White British. In the full intention-to-treat analysis, which included all participants, attendance was found to be almost identical across the 3 conditions, around 80%. Some of the women in the intervention condition, however, failed to write their plans on the questionnaire (10.6% for changing the appointment, 2.1% for travel arrangements, and 21.1% for taking time off work) or said that planning was irrelevant to them (2.4% for changing the appointment, 1.7% for travel arrangements, and 32.4% for taking time off work). A second analysis, of planning time off work, therefore, examined the 620 respondents in the intervention condition more closely and found that those who planned were significantly more likely to attend than those who did not. A 3rd analysis, excluding the 209 respondents for whom planning time off work was irrelevant, revealed that the most likely to plan were those whose initial intentions to attend were strong but whose perceptions of control over making the necessary arrangements to attend were weak. CONCLUSIONS: Two main implications of the findings are discussed: the importance of planning in implementation intentions interventions and the validity of the theoretical distinction between motivation and volition.

Speech patterning in recently admitted and chronic long-stay schizophrenic patients.

Several writers have suggested that schizophrenic patients show unusual formal, or non-content, patterns of speech. For reasons of methodology, none of their findings is easy to interpret, and important questions remain unanswered. In particular, we do not know how schizophrenic patients behave in free conversion, how their behaviour with another patient may differ from their behaviour with a psychiatrically normal partner, whether they show consistency across encounters, nor whether patients recently admitted to hospital differ from chronic long-stay patients. The first study was designed to examine the first three questions, by observing recently admitted schizophrenic patients in two free dyadic conversations, one with a schizophrenic partner and one with a psychiatrically normal partner, and comparing them with three control groups; depressive patients; patients suffering from neurotic or personality disorders; and psychiatrically normal chest patients. Consistently across their two encounters, schizophrenic subjects were found to behave similarly for the most part to all three control groups. The second study went on to examine the fourth question, by comparing recently admitted and chronic long-stay schizophrenic patients, and revealed no differences between the two groups. The investigation is discussed in the light of previous research, and it is suggested that a critical variable may be the setting and nature of the encounter in which the schizophrenic patient takes part.

The reconstruction of schizophrenic speech.

A suggestion that schizophrenic speech may be harder to understand than normal speech was tested by a technique of reconstruction. Ten schizophrenic and ten normal passages were typed onto cards, one sentence per card. Each passage was then presented with the sentences in random order, and students were asked to reconstruct what they believed was the original order. Fewer correct strings of three or more sentences were achieved for the schizophrenic material than the normal material. It is concluded that there is a detectable abnormality in the structure of schizophrenic speech, but that it stems from the relationship between sentences rather than the content of individual sentences.

Visual interaction in recently admitted and chronic long-stay schizophrenic patients.

Several reports have suggested that schizophrenic patients engage in very little Looking and eye-contact. However, previous work, much of it methodologically unsatisfactory, has been based almost always on the clinical psychiatric interview, with the result that several important questions remain unanswered. In particular, we do not know how schizophrenic patients behave in free conversation, how their behaviour with another patient may differ from their behaviour with a psychiatrically normal partner, nor even whether they show individual consistency across encounters. The first study was designed to examine these questions, by observing recently admitted schizophrenic patients in two free dyadic conversations, one with a schizophrenic partner and one with a psychiatrically normal partner, and comparing them with three control groups: depressive patients; patients suffering from neurotic or personality disorders; and psychiatrically normal chest patients. The second study went on to test whether the early descriptions of gross abnormality may be more appropriate to chronic long-stay patients than to recently admitted patients, and the design consisted of a comparison between the two groups. The first study revealed a quite unexpected pattern of results. Consistently across their two encounters, schizophrenic subjects behaved similarly for the most part to all three control groups, normal and abnormal alike. Moreover, the few differences which did emerge conflicted sharply with previous findings, including the writer's, and were no more marked in patient-patient than patient-normal encounters. The second study revealed no differences between chronic long-stay and recently admitted schizophrenic patients. It is suggested that the differences in findings between the present two studies and previous reports are most likely to be attributable to differences in verbal content: schizophrenic patients show abnormalities of visual interaction when talking about personal matters, but behave quite normally when the topic is not of immediate personal relevance.

Visual interaction and speech patterning in remitted and acute schizophrenic patients.

Previous research has suggested that schizophrenic patients sometimes show abnormal patterns of visual interaction and speech. Several important questions remain, however, and the present study was designed for two purposes; to examine the behaviour of remitted patients; and to follow up a suggestion that acute patients show abnormalities only in personal encounters. Three groups of subjects, remitted schizophrenic patients, acute schizophrenic patients and a control group of psychiatrically normal chest patients, took part, and each subject held two conversations with a nurse, one about a 'personal' topic and one about a 'neutral' topic. Unexpectedly, the remitted and acute groups were both found to behave normally, in both conversations. Taken together with previous evidence, the findings thus confirmed that acute schizophrenic patients respond situation-specifically and sometimes behave normally, but no further interpretation of the data was possible since the manipulation of topic did not have the expected effects.

Visual interaction in schizophrenic patients: the timing of looks.

Previous research has suggested that schizophrenic patients engage in very little Looking and eye-contact, at least in "personal" conversations. As yet, however, we know nothing about the precise timing of their Looks, even though this has important clinical implications. Data from three studies are presented and no evidence of abnormality is found. It is concluded that, taken together, findings from visual interaction suggest not that schizophrenic patients behave in a socially unskilled way as has sometimes been argued, but simply that they are embarassed by "personal" conversations and, like perfectly healthy people who are embarrassed, respond by averting the gaze.

Language in schizophrenia. The structure of monologues and conversations.

Experimental research into language in schizophrenia has been guided traditionally by two main assumptions: that language disturbance is widespread among schizophrenic patients and easy to detect and measure, and that schizophrenia is fundamentally a cognitive disorder in which language disturbance is part of an inability or failure to regulate one's thoughts. However, recent findings have challenged both assumptions. Two experiments are reported here, the first based on monologues, the second on conversations, which were subjected to reconstruction and discourse analyses. Schizophrenic material is found to be harder to follow than normal, and is characterised by poor reference networks and inappropriate use of questions. While some of the results are specific to the schizophrenic group, others are found also in affective patients, but none is the product of formal thought disorder. The central problem lies less in cognition than in the social process of taking the role of the other.

First diagnosis of severe mental and physical disability: a study of doctor-parent communication.

This study investigates factors associated with satisfaction with medical communication at the time of diagnosis of a child's severe mental or physical disability. Two competing social psychological models that have been used to account for parent patient satisfaction are tested: Korsch's (1968) affective model and Ley's (1977) cognitive model. One hundred and sixty six mothers of children with severe learning disabilities were questioned about the time when their child's disability was first diagnosed. Measures included the child's age when diagnosis was made, the length of time parents had been anxious, and how satisfied they were with the way they were told the diagnosis. They were asked to rate the doctor's affective behaviour and their understanding and memory of the information they received. Fifty-eight per cent of parents reported dissatisfaction with the communication. Stepwise multiple regression analysis showed that Korsch's affective scale was much the strongest predictor, entering the equation first, and explaining 35.7% of the variance in satisfaction. Ley's cognitive scale entered second, explaining a further 1.4% of the variance. Further analysis including the other predictors found that six predictors explained 40% of the variance. The implications of the findings for the training of medical staff are discussed.