In what concerns the model’s application in a threat analysis context, the threat is the undesirable behaviour, while the three components of the TPB (attitudes, social norms and perceived behavioural control) act as indicators (threat indicators = A (attitudes)+SN (social norms) +PBC (perceived behavioural control). It should be mentioned that in no way, should the model be employed for automated decision-making, especially due to its lack of comprehensiveness and high level of uncertainty (explained in more detail in the section dedicated to its limitations). The model should be employed to inform practitioners’ and policy-makers’ professional judgement by providing them with some insight as to the roots of some behaviour and informing their decisions in what concerns interaction and communication with migrants.
Perception Model Methodology
The perception model was built in three stages:
- Create an index of behaviours – An inventory of behaviours with the potential to be ‘threats’ has been identified from discussions with practitioners in MIRROR and literature review. The criteria for selecting these behaviours have been two: their impact on the security of the state and on human security in the case of migrants and the degree to which they fit in the constraints of the TPB.
- Identify the attitudes, social norms and perceived behavioural control associated with these behaviours – An extensive literature review has been carried out in order to identify evidence of attitudes/social norms that are linked to that set of behaviours.
- Create a framework of analysis modelled on the TPB – Create a framework that links behaviours with attitudes (individual beliefs about the behaviour), social norms (community beliefs about the behaviour) and perceived behavioural control (control beliefs); By starting either from an identified behaviour or and identified set of beliefs the user gain some insights into the potential correlations that exist between those beliefs and particular behaviour. This can be used either to anticipate the occurrence of certain behaviour and/or to better understand the reasons why an individual/community displayed this particular behaviour.
In addition to this two other data sets were created:
- Synonyms – These are related to the precision requirement of the perception model, namely that information on the beliefs needs to be very precise (e.g. while we can investigate general attitudes towards vaccines, the conclusions will be limited; in order to gain more accuracy, we would have to investigate the attitudes towards a specific vaccine, for example, the mRNA COVID-19 vaccine).
- Concept Associations – The concept associations dataset, also created on the basis of literature review, can act as a simple guide to better understand the type of positive or negative meaning associated with a certain topic, which is an important prerequisite when considering attitudes and social norms.
Using the model
The table below provides some practical examples of how the perception model can be applied. In the first column we find the sets of behaviours identified as being particularly relevant. These are grouped in accordance to the topic (e.g. health-related behaviours; biometrics-related behaviours). The second and third column includes the attitudes towards the behaviour and the behavioural beliefs . The fourth and fifth column include the social norms and the normative beliefs, while the sixth and seventh columns are dedicated to the perceived behavioural control and control beliefs. The last two columns as explained in the previous sections include the synonyms and concept associations (which are not part of the model per se but can be employed in order to retrieve and interpret information necessary for applying the model).
 Because this refers to individual beliefs it is very difficult to find sufficient information related to this area. However, there are individual characteristics, which can provide some insight and which have been included in the table.