Causality Orientations Theory

This is part of my “Self-Determination Theory: Condensed” series, based on Chapter 9 of Deci & Ryan’s “Self-Determination Theory”. See this page for background, and links to other articles in this series.

As presented in the previous chapter, in the presence of support for autonomy, competence and relatedness, the organismic integration process will result in optimal development. However, where basic needs supports are missing, this will result in diminished autonomous motivation, and various other forms of psychological ill-being.

Over time, persistent differences in need support can lead to significant differences in how people orient towards their environment.

Two important kinds of individual difference emerge

The first (as described in the previous chapter) is the set of regulatory styles that become dominant.
The second is the person’s causality orientation, which is the characteristic ways a person develops of orienting towards different motivational aspects of a situation: do they respond most readily to the autonomous, controlling, or amotivating aspects?

These causality orientations are the focus of this chapter.

General Causality Orientations

in 1985, Deci and Ryan proposed the existence of three distinct causality orientations:

The autonomy orientation. This is the degree to which people experience choice in their actions, reactions and opportunities in the world. People with high autonomy orientation tend to have high levels of intrinsic motivation, and identified and integrated styles of extrinsic motivation.
The controlled orientation. This is the degree to which people are conscious of external contingencies and controls, and what other people think. People with high controlled orientation tend to have low levels of intrinsic motivation, and external and introjected styles of extrinsic motivation.
The impersonal orientation. This is the degree to which people orient towards obstacles, barriers and their own incompetence. People with high impersonal orientation tend to be passive, and easily overwhelmed by their environment, or their own drives and emotions..

A person’s causality orientation moderates many of the influences on motivation that we have discussed in preceding chapters.

For example, a 2011 study found that, for people high in autonomy orientation, tangible rewards do not diminish intrinsic motivation in the way that they do in the general population (as discussed in chapter 6), whereas the diminishing effect was found to be strengthened for people who were high in controlled orientation.

The above is formally presented as:

COT Proposition I: People have three different motivational orientations - called causality orientations - that represent global-level individual differences. Causality orientations are propensities to focus on certain aspects of environments and inner capacities that concern motivation and the causes of behaviour. These are labelled the autonomy orientation, the controlled orientation and the impersonal orientation. These orientations affect people’s situation-specific motivation, as well as their general need satisfaction, behaviour, and experience.

COT does not suggest that people have one exclusive orientation. Rather, each person has some degree of each of the three orientations. The General Causality Orientations Scale provides people with a score for each orientation.

In general, these scores are considered to be general, or cross-domain - but for some people, their causality orientation varies between contexts (e.g. between sports events vs. in school). It has also been shown that particular causality orientations can be primed by various methods. Both these considerations are discussed later.

Why Study Individual Differences?

The main reason for studying individual differences is to allow SDT to account for significant individual deviations from the general predictions of SDT.

People in controlling contexts are not always controlled: some continue to be autonomous. And similarly, some people in autonomy-supporting contexts continue to operate as if they are controlled. While mini-theories such as CET and OIT are highly predictive at a group level, they do not account for these outliers - and this is the role played by Causality Orientations Theory within SDT.

There are at least three ways in which individual differences can affect outcomes:

seeking out - e.g. people with a high autonomous orientation are more likely to seek out situations that are autonomy-supportive
interpretation - different causality orientations may lead two people to interpret the same event in different ways - one may experience it as controlling, when the other does not.
influencing social contexts - e.g. someone with a strong controlled orientation may “pull on” their teacher or supervisor to control them, leading to that person being more controlling than they might have been otherwise.

COT includes descriptions of each of these effects.

Development and Causality Orientations

Causality orientations are both an output of past development, and an input into future development. Through the mechanisms just described, each causality orientation tends to feed into a feedback loop, further strengthening that orientation.

How causality orientations develop is described by the following formal proposition.

COT Proposition II: Causality orientations are developmental outcomes that are influenced over time by biological and social-contextual factors that impact satisfaction of the basic psychological needs for autonomy, competence and relatedness. To the degree that individuals’ social environments are substantially and persistently autonomy-supportive, controlling, or amotivating, people will, respectively, tend to develop strong autonomy orientations, controlled orientations or impersonal orientations.

Assessing General Causality Orientations

The General Causality Orientations Scale (GCOS) provides 3 sub-scores, one for each causality orientation, based on a series of questions.

Causality orientations have been shown to be empirically distinct from the “Big Five” personality traits (openness, conscientiousness, extraversion, agreeableness and neuroticism), though there were some correlations:

Autonomy was positively correlated with extraversion and openness
Control was negatively correlated with agreeableness
Impersonal was positively correlated with neuroticism.

Studies have also shown that incorporating causality orientations into models alongside the “Big Five” personality traits increases predictive power of these models.

Autonomy Orientation: Empirical Results

Deci & Ryan’s 1985 study showed that the autonomy orientation was positively correlated with self-esteem, ego development and self-actualization.

Further studies have shown that autonomy orientation is associated with a wide range of other variables that represent psychological well-being, persistence, complex and flexible thinking, prosocial involvement, healthy behaviour change, and effective interactions with social partners.

Highlighting a few specific studies:

A 1994 study showed that high autonomy orientation was associated with students adopting learning goals, rather than performance goals, and having higher confidence in their academic abilities.
A 1996 study found that autonomy orientation was predictive for successful and sustained weight loss among a group of morbidly obese patients.
A 2002 study found that high autonomy orientation had multiple positive effects on the quality of romantic relationships.
A 2009 study showed that people with high autonomy orientation coped better with negative emotions arising from exposure to disturbing films.
A 2003 study of prosocial behaviour among community volunteers found that autonomy orientation predicted overall levels of engagement and participation.
A 2003 study showed that autonomy orientation protected people from negative feelings resulting from comparisons with others.
A 1983 study of mothers with 1-year-olds showed that autonomy orientation of mothers predicted secure attachment on the part of their infants.
Other studies have shown that those with high autonomy orientation experienced low levels of boredom, were more likely to focus on interest and challenge at work, and were also more likely to be autonomy-supportive of others.

Controlled Orientation: Empirical Results

Deci & Ryan’s 1985 study showed that the control orientation was positively correlated with public self-consciousness, and also with a hard-working “Type-A coronary prone behaviour pattern” (i.e. people likely to work themselves into a heart attack!).

Other studies have shown that people with a high control orientation tend to be more Machiavellian, tend to cheat more when they have the opportunity to do so, preferred image-based rather than quality-based advertising, and placed a very high value on amassing wealth.

Among adolescents, the controlled orientation is associated with less flexibility and more rigidity. It’s also associated with very firm commitment to goals, and with conscientiousness (which can appear to be positive, but also risks harming that person by thwarting their need for autonomy).

The controlled orientation has also been associated with a number of other problematic effects.

A 1994 study of students found that those high in the controlled orientation tended to adopt performance goals rather than learning goals, and demonstrated ego-involved persistence in the face of failure (which, as we saw in chapter 7, is a non-optimal form of persistence).
A 2004 study found that strong controlled orientation predicted gambling behaviour, specifically greater frequency, larger sums involved, more negative gambling consequences, and greater proneness to disordered gambling. It also predicted heavier drinking of alcohol, and more alcohol related problems.
A 2001 study linked the controlled orientation to a series of problematic driving behaviours: feeling more angry about other drivers’ behaviour, more aggressive driving, and more traffic citations.
More generally, there seem to be links between the controlled orientation and tendencies towards interpersonal aggression. This is consistent with the view that controlled people feel more like objects, reducing their capacity for empathy.

Impersonal Orientation: Empirical Results

Deci & Ryan’s 1985 study showed that the impersonal orientation was related to social anxiety, public self-consciousness, self-derogation, depressive symptoms, lower self-esteem and lower ego development.

Later studies have shown a number of other effects:

A 1999 study found that impersonal orientation was related to the powerlessness and self-estrangement aspect of alienation.
The same 1994 study mentioned above found that students with high impersonal orientation tended to hold performance goals (rather than learning goals), but also had low confidence in their ability to do well.
A 2015 study showed that people with schizophrenia had lower autonomy orientations, and higher impersonal orientations.
In a 2005 study of adolescents, high impersonal orientation was associated with procrastination, denial of internal conflicts, putting off decisions and actions, maladaptive coping styles, and neuroticism.
A 1987 study showed that women diagnosed with anorexia had higher impersonal orientation than a control group with the same demographics.

Collectively, these empirical results led to a 3rd formal proposition of COT:

COT Proposition III: Causality orientations affect people’s effectiveness in engaging with their surroundings, as well as their psychological well-being, as mediated by types of domain- and situation-specific motivations and need satisfactions. The autonomy orientation promotes greater integration of personality, which strengthens itself and promotes effective performance and well-being. The controlled orientation promotes introjection and rigidity, which strengthens itself and promotes less-effective self-regulation and less positive experience. The impersonal orientation promotes the experience of ineffectiveness and amotivation, thereby strengthening itself and leading to the least effective performance and lower well-being outcomes.

Locus of Control and Locus of Causality

Deci and Ryan’s 1985 research also found that the impersonal orientation was related to an external locus of control, i.e. the belief that one’s behaviour cannot affect outcomes.

A person’s locus of control refers to their belief about whether they have the ability to affect outcomes with their behaviour.

This is related to, but not to be confused with perceived locus of causality (PLOC), introduced in chapter 6.

An I-PLOC and an E-PLOC both correspond to an internal locus of control, in that the person believes themselves capable to affect outcomes (the difference between an I-PLOC and an E-PLOC depends on the perceived source of the desire for that outcome).

An external locus of control corresponds to an impersonal locus of causality, that is a feeling that one cannot attain outcomes, a feeling of passivity and a sense of amotivation.

Other Theoretically Important Empirical Results

The GCOS has been used to confirm a number of important theoretical points that are central to SDT.

That autonomy is a reflection of greater integration in personality
That causality orientations relate to need satisfaction, with higher autonomy orientation predicting greater need satisfaction
That causality orientations lead to individual differences in terms of openness vs. defensiveness, and quality of social interactions.
That causality orientations explain variation in regulatory styles.

We now look at these points in more detail, in turn.

Autonomy and Integration

We saw in chapter 8 that autonomy is a key support for the natural process of integration, and that integration supports acting with greater feelings of autonomy.

A series of 1992 studies looked at the relationship between autonomy and integration in terms of differences between individuals. They divided the individuals in the study into groups based on their dominant causality orientation. They then looked at various aspects of integration or consistency between the groups.

One such aspect was to compare the free-choice behavioural measure vs. the self-reported interest measure of intrinsic motivation. Across the entire group, the average correlation was 0.35, which matched the level of correlation between these measures that was seen in the 1999 meta-analysis we discussed in chapter 6. However, within the autonomy-oriented sub-group, the correlation was over 0.6, while within the control-oriented sub-group, the correlation was effectively zero. This suggests that the free-choice behaviour of the autonomous-oriented group was for the most part genuinely driven by intrinsic motivation, while the control-oriented group’s free-choice behaviour was likely driven by controlling thoughts or introjected feelings.

A second aspect was the relation between traits (as assessed by questionnaire) and behaviours. They looked at the correlation between conscientiousness, as assessed as part of a personality test, and actual conscientious behaviour. Again, the correlation was much stronger for the autonomy-oriented group, than it was for the control-oriented group.

The final study in this series compared how people perceived themselves with how they were perceived by others. This correlation was also somewhat stronger for the autonomy-oriented group, than it was for the control-oriented group.

A 2011 study looked at integration in a very different way. Participants were presented with various experiences, positive and negative, from three years earlier, and asked whether they accepted those experiences as part of who they were today (at the time of the study). Those who were classified as autonomy-oriented were accepting of all experiences, positive and negative. But those who were classified as control-oriented tended to accept only the positive experiences, and not the negative ones.

This study therefore showed greater integration among those with a more autonomous orientation, and evidence of compartmentalization among those with a more controlled orientation.

This study also looked at priming, and found that priming participants for autonomy also led to them acting like the autonomy-oriented group (we shall discuss priming in more detail shortly).

Causality Orientations and Need Satisfaction

SDT predicts that people who are more autonomous will (1) have a clearer awareness of their basic needs, and (2) be more effective in taking actions to meet those needs. This implies that more autonomous people should have higher levels of basic needs satisfaction. And as further consequences of this, they should perform better, and have greater well-being.

A 2004 study in two workplaces tested this hypothesis, analysing workplace performance data, together with the GCOS, needs satisfaction data, and a questionnaire about how participants perceived their work climate.

The following results emerged:

need satisfaction predicted both workplace performance, and well-being
perceived autonomy support in the workplace environment predicted need satisfaction.
individual differences in autonomy orientation predicted all of need satisfaction, performance and well-being, independently of the level of autonomy support perceived in the workplace.

A further 2008 study also showed that autonomy orientation also predicted work motivation, job satisfaction, and individuals’ commitment to their work.

Studies in 1996 and 1998 using the GCOS showed that people high in autonomy and low in control were less likely to take undue credit for success, and to duck responsibility for failures. They were also less likely to use self-handicapping strategies and avoidant approaches to coping during stressful periods.

A 1996 study in relationships between students and their parents found that students high in autonomy and low in control on the GCOS were more honest with their parents, enjoyed interactions with their parents, and felt better about those interactions. A follow-up study found similar effects in students’ other relationships (e.g. with their peers).

A 2003 study found that people with high autonomy tended to accept greater responsibility, and be less blaming, with regard to interpersonal problems. Various other studies have shown people with high autonomy to use fewer lies in explaining wrongdoings, and provide more apologies, especially more complex apologies where they had hurt others.

These results are important because they suggest that autonomy can provide strong support for relatedness. This is contrary to popular theories (e.g. from Gilligan, 1982 and Jordan, 1991) that autonomy and relatedness are at odds within relationships.

This sort of misconception likely stems from an interpretation that autonomy is principally about independence, rather than volition - a misconception that merits some detailed comments…

Reflective and Reactive forms of Autonomy

Several theorists of personality, including Murray (1938) and Gough and Heilbrun (1983) have understood autonomy in terms of independence from others.

In 1996, Koestner and Loesier coined the term reactive autonomy to refer to this concept, and drew a contrast with reflective autonomy which is understood to mean acting with a sense of choice, and is aligned with the meaning of autonomy within SDT.

Reactive autonomy had previously been related to a dislike of teamwork, being more likely to drop out of college, and engagement in criminal behaviour.

Kostener and Loesier showed that reactive autonomy was uncorrelated with reflective autonomy (i.e. the GCOS sub-scale of autonomous orientation), but was positively correlated with the GCOS sub-scale of controlled orientation. Furthermore, reactive autonomy predicted more negative events and negative moods, while reflective autonomy predicted more positive events and positive moods.

Further results included the observation that:

reactive autonomy was associated with negative interactions with authority figures, while reflective autonomy was not
reflective autonomy was associated with positive interactions with peers, while reactive autonomy was not.
those high in reflective autonomy tended to follow the advice of credible experts, while those high in reactive autonomy were significantly less likely to follow such advice.

It is clear from the above that reflective autonomy (what SDT refers to as autonomy) and reactive autonomy (which could also be called independence) are very different concepts.

Autonomy as defined in SDT results in openness, positive engagement with peers, and a lack of defensiveness. Independence often results in more negative relationships with peers, and rebellious reactions to authority figures.

Causality Orientations and Self-Regulation

Various studies have examined the hypothesis that people with a high autonomy orientation in GCOS tend to be more autonomous in their regulation of behaviours (in the sense discussed in chapter 8).

A 1996 study of medical students taking an interviewing course found that high autonomy orientation did indeed predict more autonomous regulation. The presence of an autonomy-supportive climate also led to more autonomous regulation (as predicted by OIT), but the effects of a high autonomy orientation could be seen independently of this effect.

A 1996 study of obese participants in a 6 month weight loss program also showed that high autonomy orientation predicted both reliable attendance and maintained weight loss over a 2 year period.

The Hierarchical Model

So far, we have looked at differences in motivation between individuals, but we have not yet considered variations in motivation that undoubtedly occur within an individual, in different contexts.

In 1997 Vallerand proposed a hierarchical model of motivation, in which motivations are grouped into three categories:

autonomous motivation (i.e. intrinsic motivation or identified/integrated extrinsic motivation)
controlled motivation (i.e. introjected or externally regulated extrinsic motivation)
amotivation

He proposed assessing these at three different levels of generality

at a global level for the individual (i.e. across all domains)
within a specific domain, such as academics, work or sport
within specific situations.

Vallerand proposed that motivation at a particular level is determined by social-contextual factors at that level, in combination with the motivation at the next level in the hierarchy - and further that it is the motivational style at a particular level in the hierarchy that is best used to predict outcomes that arise from motivation e.g. quality of performance, feelings (positive or negative), or even physical symptoms.

It is worth noting that from a developmental perspective, motivation can also flow upwards through the hierarchy. For example, an individual class that is autonomy-supporting can feed into greater autonomous motivation in academic work, which can in turn feed a more generalized autonomous motivation.

Priming Motivational Orientations

An individual’s general causality orientation can be seen as a kind of “average” across all situations and domains, but it is normal for different orientations to come to prominence in different contexts.

This is covered by a fourth formal proposition of Causality Orientation Theory:

COT Proposition IV: All individuals have three causality orientations to some degree. Subtle cues in the environment may make different orientations more salient at that time and place. Thus, it is possible to prime people’s motivational orientations such that their behaviour and experience will be significantly affected by the primed motivation, even if that orientation is, in general, relatively weak.

As mentioned earlier, a 2011 study found that the effects of primed causality orientations could be just as strong as the effects of individual’s general causality orientations.

A standard priming procedure is the scrambled sentence method. Participants undertake a 5 minute task where they have to rearrange sentences, whose word order has been scrambled, into the correct order. The autonomous orientation can be primed by including in the sentences words such as interested, challenge, satisfied and mastering. The controlled orientation can be primed by including in the sentences words such as obligation, evaluated, constrained, demanded and forced.

A 2003 study involving a puzzle task showed that such priming could affect not only participants’ feelings about the task (their level of interest and feeling of choice), but also their actual performance on the task. The effect of the priming was as strong as the effect of different GCOS orientations, or the effect of autonomy-supporting vs. controlling task environment.

A follow-up study revealed two further interesting results:

One was that the effect of priming was stronger for those who did not already have a strong underlying orientation of the type being primed. i.e. priming could boost a low causality orientation, but did not substantially boost a high one.
The second was that the positive benefits of an underlying high autonomous orientation on autonomous motivation were still visible, independently of the priming.

Further studies of this kind of scrambled-sentence priming have shown that several other effects previously observed with GCOS can also be generated with priming:

A 2006 series of studies showed that priming could influence defensiveness, self-serving bias, and self-handicapping. In each case, the results matched results that had previously been shown using GCOS scores, with autonomous primes resulting in the lowest levels of these negative behaviours, and impersonal primes the highest levels.
Further studies have shown that these kinds of primes are effective in modifying behaviour in stressful interview situations, and when making speeches.

These priming studies are important for a couple of reasons:

They provide strong support for a causal interpretation of the correlations observed between causality orientations and various observed behaviours, feelings etc.
They show that causality orientations can operate automatically (i.e. sub-consciously).

Autonomy and Quality of Relationships

A 2010 lab study looked at the effects of factors known to be autonomy-supportive, or controlling, upon the development of new relationships. Participants were introduced to each other, and invited to share personal information with each other in an environment that had been constructed to be either autonomy-supporting or controlling (or a neutral control environment).

The study found that autonomy-supportive factors led to greater relationship satisfaction, more positive emotions, and well-being, while controlling factors had the opposite effects.

This suggests that autonomy and control can significantly affect not only motivation and performance, but also the quality of interpersonal relationships.

Concluding Comments

Generalized Causality Orientations reflect individuals’ tendency for their motivation to be autonomous or controlled, or to be amotivated.

These individual differences explain why different people respond differently within the same, or similar, social contexts.

The autonomy orientation is strongly associated with the healthiest development and greatest personality integration, the controlled orientation is associated with more rigidity and defensiveness, while the impersonal orientation is associated with amotivation and the lowest well-being.

Key Concepts from this Chapter

Autonomy orientation: The degree to which people experience choice in their actions, reactions and opportunities in the world.

Controlled orientation: The degree to which people are conscious of external contingencies and controls, and what other people think.

Impersonal orientation: The degree to which people orient towards obstacles, barriers and their own incompetence.

COT Proposition I: The three causality orientations are: autonomy, controlled and impersonal.

COT Proposition II: Causality orientations develop based on the support or thwarting of basic psychological needs. Autonomy-supportive, controlling, or amotivating environments each support growth of the corresponding causality orientation.

General Causality Orientations Scale (GCOS): A measure of an individual’s overall orientation in each of the three causality orientations.

COT Proposition III: Each causality orientation promotes a particular motivational style (autonomous, controlled or amotivation), and impacts performance and well-being, with the autonomy orientation having the most beneficial effects and the impersonal orientation the most detrimental effects.

Locus of Control: A person with an internal locus of control believes that they can affect outcomes through their behaviour. A person with an external locus of control does not have this belief. Not to be confused with Locus of Causality (see chapter 6).

Reactive Autonomy: Roughly synonymous with independence. The word autonomy has been used in psychology literature to refer to this concept, but this is a distinct concept from the way the word autonomy as used in SDT, and has shown to have very different (and often negative) effects, especially with regard to relatedness.

Reflective Autonomy: This term refers to autonomy in the sense used in SDT, used in contexts where there is a need to distinguish from reactive autonomy

COT Proposition IV: In a particular time and place, which causality orientation is most salient in an individual, will be affected by the environment. In particular, it is possible to prime particular motivational orientations.

Priming: A process by which a particular attitude or idea is sub-consciously made salient for a participant in an experiment.

Scrambled sentence method: A priming procedure based on an exercise in which the words of sentences are rearranged. The choices of vocabulary in the sentences have been shown to result in priming of various different causality orientations.