2.2 Theory of Reasoned Action (TRA)
2.2.5 Applicability and criticisms of TRA
Since its introduction, TRA has been extensively used in several empirical studies to explain behaviour (Prestholdt, Lane & Mathews, 1987; Fredricks & Dosset, 1983;
Ngo, 2013). There has also been empirical credence reported on the strong relation between attitudes, subjective norms and behaviour (Sheppard, Hartwick &
Warshaw, 1988; Quine & Rubin, 1997; Stone, Jawahar & Kisamore, 2010). In many cases, a person's attitude and subjective norms toward a behaviour have been presented as sufficient determinants of their performance of the behaviour (Bowman & Fishbein, 1978).
Positive attitude and social norm indicate a strong likelihood of performing the behaviour (Glanz, Lewis & Rimer, 1997). Several other analyses have shown that TRA is a useful model to predict behaviour (Robinson & Doverspike, 2006). This is
also strengthened by the assumptions and methodological approaches of TRA which have been viewed as more constrained and present a less flawed approach to explaining behaviour. This led to the conclusion that "the model yields stable predictions" of behaviour (Ryan & Bonfield, 1975: 125; Sheppard, Hartwick, &
Warshaw, 1988; Kim & Hunter, 1993).
Despite its credence and success in empirical applications, TRA has also faced criticism focused on some of its main tenets and the completeness of its components.
While several studies have reported significant relationships between attitude, subjective norm and behaviour (Sheppard, Hartwick, & Warshaw, 1988; Bagozzi &
Warshaw, 1992; Albarracin, Johnson & Fishbein, 2001; De Boer, 2003), there are other studies that challenge the validity of these relationships (Wicker, 1969; Eagly
& Chaiken, 1993). There have been suggestions that more constructs need to be added to strengthen the explanatory power of TRA (Fishbein & Ajzen, 2010). One of the criticisms has been on the relationship between intentions and behaviour, a key tenet of TRA. While there have been many studies supporting the existence of a strong relationship between intentions and behaviour (Sheppard, Hartwick, &
Warshaw, 1988; De Boer, 2003; Fishbein & Stasson, 1990), others have suggested that the relationship between intentions and behaviour is rather weak (McQuarrie &
Langmeyer, 1987) and may not sufficiently explain behaviour. Attitudes and subjective norms are also shown to be related to intentions and to behaviour but the exact nature and strength of these relations are still uncertain (Gooding, 1994).
It has also been argued that in explaining behaviour, TRA sees a person through his rationality and not his humanity (Heap et.al, 1992). This assumption that human beings are rational, systematically utilizing and processing information when deciding what action to take has been criticized as ignoring human limitations on information processing and handling that would make them rational. Though TRA assumes that people can cope with much information, there have been arguments
that there are limits to how much information one can handle. Ajzen and Fishbein (1980) concurred with this but argued that it does not undermine the logic of TRA.
There have been other arguments that though TRA views human beings as rational, this is only to some extent: their rationality is bounded. From systematic observation in organizations, scant evidence of behaviour based on rationality emerges (Jones 1985). This was also based on the notion that TRA ignores the fact that intentions, ends and preferences, which are core in rational behaviour, are neither givens nor stable and as such, may change (Elster, 1987). Further, for it to hold, it requires complete knowledge and anticipation of the consequences of behaviours which is always hard to attain. In view of these, Simon (1982) argued for ‘bounded rationality’: individuals behave rationally, but within some bounds which compel them to settle for satisficing rather than optimizing (Heap et al. 1992) their preferences or outcomes of their behaviours. Bounded rationality thus argues that though humans are rational, goal oriented and adaptive, they sometimes fail in making decisions due to their cognitive and emotional architecture. Because perfect knowledge never exists, it means that all choices imply risk. Human behaviour is quite complex to be understood from this perspective alone. However, Fishbein &
Ajzen (2010: 24) explain that the term reasoned action does not mean that they consider actors to be rational and to deliberate at length before engaging in any behaviour. According to them, the theoretical framework does not assume rationality but encompass both deliberate and spontaneous decision making.
Related to the above, the other challenge with the theory has come from its reliance on cognitive structures to explain behaviour. Such reliance would mean that the theory does not recognise connections between individuals within the social contexts in which they act. It argues that though the theory recognises social norms, their measurement is confined to a limited consideration of individual perceptions of these social phenomena and thus making it difficult to predict how individuals are likely to behave (Vallerand, Deshaies,Cuerrier, Pelletier, & Mongeau, 1992). A
further criticism has been that TRA assumes that when one forms the intention to act, they will act without any limitation. This might at times not hold. Even if attitudes and subjective norms are positive, other constraints can prevent behaviour (Fishbein & Ajzen 2010:64; Cheng & Ho, 2001). TRA can adequately predict behaviours that are relatively straight forward and under volitional control within circumstances where there are no other constraints to action. But not all behaviours are under volitional control. At times constraints, both personal and environmental could constrain execution of behaviour even if one has positive intentions towards the behaviour (Cheng & Ho, 2001; Hungerford & Volk, 1990).
These criticisms have led to the need for extension of TRA to add more predictors of behaviour, especially for non-volitional behaviours instead of just relying on attitudes and social norms which could mainly apply in explaining volitional behaviours where individuals have complete control. There are additional variables to further constrain the model to strengthen its power to explain behaviour. The next section discusses such an ‘extended TRA model’: the Theory of Planned Behaviour (TPB).