2 MARCO TEÓRICO
2.6 SUBSISTEMAS DE LA GESTIÓN DEL TALENTO HUMANO
2.6.4 DESARROLLO DE PERSONAS
3.4.1 Previous comparison studies of theory of planned behaviour and health belief model
Despite the extensive use of both models in the field of health psychology few studies have been explicitly compared the two models, although there are a number that have combined the two models. One of the earliest comparisons of the HBM and TRA (Oliver & Berger 1979) demonstrated that the TRA explained 50% of the variance in
Chapter 3: Theoretical Basis of Thesis intent to have a flu vaccination, compared with the 30-35% explained by the HBM. A study exploring different health behaviours (Mullen, Hersey & Iverson 1987) found that the theories explained different relative proportions depending on the behaviour examined. After accounting for demographics and baseline behaviour, the TRA and the HBM explained approximately equal amounts of variance in measurement of the number of cigarettes smoked (14.7%) and fried food consumption (13.8% and 13.6% respectively). The TRA outperformed the HBM on one behaviour (attempts to quit smoking) explaining 24.7% and 7.9% of the variance respectively. On two behaviours the HBM explained more variance than the TRA - exercise (17.9%/ 9.5%) and sweet food consumption (11.5% / 9.6%).
In a study of medication compliance in women with urinary tract infections (UTIs) (Ried & Christensen 1988), the HBM alone only explained 10% of the variance in compliance, whilst the TRA explained 23% of the variance. When the two theories were combined the resultant model explained 29% of the variance in medication compliance.
In an assessment of the comparative power of the TPB and the HBM (Conner & Norman 1994) to explain screening intentions and behaviours, the HBM and the TPB explained 55% and 52% of the variance in intention respectively. The HBM however contained a larger number of variables (7 compared to 5), thus the R^ for this model would be expected to be higher. When the two models were combined the proportion of variance explained increased (61%) but non-signifrcantly. The significant predictors in this combined model were benefits, barriers, health value, behavioural beliefs and attitudes. Neither model explained much of the variance in behaviour, and when combined only 5% of the variance was explained.
An exploration of the TRA and HBM (Vanlandingham, Suprasert, Grandjean & Sittitrai 1995) found that components of both models were associated with prior condom use when visiting prostitutes in Northern Thailand. In direct comparisons of the models the TRA was found to be the superior theory, it also classified a higher proportion of the cases correctly. The effect of peers (subjective norm) was found to be the largest influence in this model on behaviour, as measured retrospectively.
Chapter 3: Theoretical Basis of Thesis The two models have also been combined in studies to produce emergent models that contain components of both. In exploring adherence to different malaria prophylactics (mefloquine c.f. chloroquine and proguanil) on return from foreign travel (Abraham, Chfr & Grabowski 1999) a combined model was used. Intention, PBC, attitude, injunctive norms, perceived side effects, perceived susceptibility and perceived severity explained between 39.5% and 50.2% of behaviour, with adherence in the affected region explaining an additional 10.1% of variance for the latter regime (chloroquine and proguanil). In the explanation of intention the combined model performed very well, explaining 65% and 77.1% of the variance in intention for the two drug regimes. The main correlates of intent were perceived behavioural control, perceived side effects, perceived susceptibility and perceived severity. Whilst not providing a direct comparison, this study has provided evidence for the additive value of the HBM to the TPB.
The previous studies have therefore produced somewhat inconclusive results, with the TPB/TRA usually performing marginally better in most studies than the HBM. In studies that have combined the two models, the combined model explains more variance in intention, indicating that there may be components of each model which explain unique variance.
3.4.2 Differences and Commonalities in the Models
The theory of planned behaviour and the health belief model have both been used extensively in the health psychology literature. It is clear from using combined models that there is both shared and unique variance between the models. Both models investigate the person’s own evaluations of genetic testing. In the HBM this is operationalised as the perceived benefits and barriers towards them undergoing genetic testing, and in the TPB attitudes are explored - i.e. their overall evaluation of the behaviour.
The influence of attitude, or the balance of benefits and barriers, may be important factors for a person in deciding whether to undergo testing. Genetic screening remains an emotive issue, particularly in connection with prenatal testing, although many people do not fully understand the issues involved. Those who are aware of a particular disease in their family may be better informed, however, without a clear explanation
Chapter 3: Theoretical Basis of Thesis they may still not grasp the implications of testing, or may hold misinterpretations of the information that they have received. Without fully understanding the issues involved it might be difficult to form a strong opinion about testing based on fact. This raises the possibility that decision making may be based less on concrete factors such as benefits and barriers, but more on general measures of attitudes, and emotional evaluations such as anticipated affect. There are clearly differences between the benefits and barriers in the HBM and attitudes in the TPB, however it is at this point that the two theories are most closely related, and it is expected that shared variance would occur when these three variables are all in the model.
At this point the models diverge. The TPB focuses on the person’s perception of what other people would want them to do (subjective norm), and whether they believe that they can actually carry out the behaviour (perceived behavioural control). Perceived behavioural control is hypothesised to influence not only the intention to undergo testing, but also whether the person actually undergoes testing. These variables are not adequately measured in the health belief model. The desire to please others may be construed as a particular benefit or barrier to undergoing an activity, but is not explicitly included in the model. Perceived behavioural control is not measured in the health belief model, and may be particularly important in the decision to have a genetic test that is not under volitional control. Perceived behavioural control could be a specific example of a barrier to testing, but again is not explicitly tested in the HBM for which there is no standardised format beyond the main variables. These two parts of the theory of planned behaviour would therefore be expected to continue to explain unique variance when the HBM is entered into a model.
The HBM introduces two concepts that are not included in the TPB - perceived susceptibility to the disease and the perceived seriousness of the disease. Perceived susceptibility might also play a role in the decision to undergo testing, as a person who does not perceive themselves to be susceptible to a rare disease in theory should be unlikely to request testing for it. This has been supported in research reported so far, as intention is correlated with perceived susceptibility, however even when informed that their risk of hereditary colon cancer was very low, many people in the general population persist in their desire to have a genetic test (Graham et al 1998). The role of perceived severity of the disease, especially with regard to cancer, as discussed
Chapter 3: Theoretical Basis of Thesis previously may not be particularly useful, as there is oflen little variation in this predictor variable. Although not explicitly explored within the TPB, these factors may contribute to the formation of a person’s attitudes, although it is most likely that perceived severity and perceived susceptibility will emerge as distinct factors explaining unique variance in intent to have a genetic test.
There are therefore commonalties between the two models, and distinct differences, so a comparison of these two models is appropriate to explore the actual influences on intentions to have a genetic test. These models cannot be regarded as all-inclusive. Clearly a general model which aims to describe the uptake of all health behaviours (as these models do), cannot be expected to cover factors unique to certain behaviours. In order to identify other important correlates of intention other measures will also be taken to explore what additional effect these might have on the variance explained by each model.