Process Decision Making Theories–the Consensus

pantheon2imagesThis post summarizes the second half of a paper entitled: “Theory-informed design of values clarification methods: A cognitive psychological perspective on patient health-related decision making,” that appeared in Social Science & Medicine 77 (2013), and was written by Arwen H. Pieterse, Marieke de Vries, Marleen Kunneman,  Anne M. Stiggelbout, and Deb Feldman-Stewart. The post includes a summary of the general agreements of the four theories and seven recommendations based on these agreements to aid value clarification for patients. Again, I think it is almost amazing that these theories are being examined in one paper.  I am impressed by the clarity and usefulness of the examination.

The identified theories generally agree that a preference emerges through stages in decision processes: an initial mental representation of the decision situation, information acquisition and interpretation, including the generation of one or more gist representations, information integration into an evaluation, and the declaration of a preference. Also, a commonality across these theories is that difficult decisions tend to be made using compensatory strategies, as was evidenced when people face decisions that are new ; when they are personally involved in decisions; or when they make decisions under uncertainty.

Fig. 1 provides a schematic overview of naïve (as opposed to expert) decision making processes over time based on the selected theories, from the point when potentially appropriate options are presented until after a preference is declared. Fig. 1 further identifies potential processes of values clarification as derived from the theories that VCM could target in order to lower the burden of decision making. For the sake of clarity the authors represent the values clarification processes in Fig. 1 as if being serial. Building on Fig. 1, they present recommendations on how to support patient values clarification.

Help Optimize Mental Representations (Representations)

A basic assumption of the theories selected is that mental representations constitute a core element of decision processes . Mental representations affect what information is attended to and what information is ignored. Parallel Constraint Satisfaction theory postulates that any presented information can be included in the representation of options regardless of its relevance. Diff Con theory suggests that salient affective, cognitive or perceptual aspects of options at the start of a decision process partly determine its course and the mental representation of the decision.  For example, stereotypes can be part of that early representation and, thus, affect decision processes without decision makers’ awareness, even in situations when they would prefer to avoid that influence if they were aware of it . Similarly, choices may be affected by information that individuals in fact identify as irrelevant to the decision. Mental representations of decisions can also be compromised because individuals do not understand available information correctly, which may especially be true for probability information . Decision makers will tend to give less weight in their decisions to attributes that they feel they do not understand well. In terms of Fuzzy-trace Theory, decision makers need to understand the essential meaning of the information for their situation which will be stored as gist representations and will in turn cue which values will be retrieved from memory. Of note, the aim of presenting information that is deemed relevant to patients represents a major challenge as clinicians do not agree on what information is critical, clinicians and patients differ in their information priorities, patients differ in their information priorities, and patients’ information needs change during the decision making process.

Include all potentially appropriate options and their attributes (Representation)

Image Theory, Diff Con theory, and Parallel Constraint Satisfaction theory all claim that we typically screen initially available options when many are offered in order to focus on fewer relevant options. Individuals may discard options early in the process without considering all relevant information, to narrow the pool of options which they will consider for final selection. This may lead to suboptimal choices. Also, as Image Theory and Diff Con theory suggest, early processing often relies on non-compensatory processes. This may result in individuals focusing on only one or a few attributes of options which may, in turn, lead them to discard options early. Image Theory postulates that our natural screening processes aim to exclude bad options rather than to include good options. Prompting decision makers to include potentially appropriate options was shown to result in a larger pool of options retained . To address this screening inclination, therefore, VCM should encourage individuals to include all potentially appropriate options in their consideration set.

Suspend the selection of an initially favored option (Pre-selection)

Once we have encouraged representation of all options, Diff Con theory and Parallel Constraint Satisfaction theory suggest that we naturally tend to select an initially favored option early in the decision making process. This strategy helps anchor incoming information to understand and integrate it into the representation more readily. Also, selecting an initially favored option at an early stage saves effort because each competitor has to be compared only with that preliminary choice. Decision makers may select an initially favored option too soon, for example before they have considered all potentially relevant information. They may select an early preference based on irrelevant attributes, for example because a significant other who is not knowledgeable but nevertheless trusted on the issue recommends it. Diff Con theory acknowledges that in this process preference may change and that the final choice need not necessarily be the same as the initially favored option. In general, both the reported values of the attributes (e.g., how often one has to go to hospital to undergo radiation therapy) and their weights (e.g., how important traveling to the hospital is for the patient) have been shown to shift to make one option dominate competitors. Considering options increasingly positively or negatively before receiving all relevant information could be counterproductive in reaching “good” decisions.

Remind patients of their array of values (Integration)

Fuzzy-trace Theory proposes that decision makers integrate their gist representations of information and knowledge with values that they retrieve from their long-term memory. Importantly, Fuzzy-trace Theory suggests that the retrieval of values is context-dependent. In effect, even strongly-held and relevant values may not be retrieved at the time of decision making. The priority of values in long-term memory and the cuing of values in the context that the decision maker faces will jointly determine the accessibility of the value at the time of decision making. VCMs should encourage patients to consider the array of values they hold in long-term memory and that are relevant to the decision. This may be done, for example, by helping them identify each value that is important to their choice under conditions of low cognitive burden, a strategy that should encourage the patient to be comprehensive.

Facilitate the weighing of attributes of options (Integration)

The selected theories share the view that complex decisions tend to be made using compensatory strategies. VCMs should facilitate the weighing of pros and cons of options, regardless of whether the actual weighing should take place deliberately or not. Weighing may be facilitated, for example, by presenting options side-by-side in table format. Simultaneous display of options facilitates attribute-wise comparisons between alternatives and thus compensatory decision strategies . The consideration of all relevant information is expected to require time. When done deliberately, it is expected to require cognitive effort, and may be emotionally demanding. Additional time and cognitive effort will mostly be limited to making tradeoffs and will not typically include information search because aids should by definition provide complete information as much as possible.

Offer time to decide (Representation, Pre-selection, Integration)

The theories reviewed suggest that we naturally restructure our mental  representations to reach a decision. Evidence suggests that the extent to which mental representations are restructured to support the initially favored option is moderated by perceived time pressure. When facing a health-related decision, often actual time pressure is not at stake and individuals can take at least days before committing to an option.  Providing time for making decisions is expected to encourage unbiased processing of information and thorough integration of pros and cons. There is little evidence regarding what specific period of time to recommend. Patients may, for example, be invited to think about their options and/or discuss them with others and to come back and see their clinician a day or a week later, before making the final decision.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.