Decision Biases
16 important questions on Decision Biases
What are the 2 types of framing?
• Positive vs. negative (attribute framing)
• Relative risks vs. absolute risks -> make a decision to present things in a proportional change (e.g. this will increase your risk by 1/3; this will double your chances of a stroke) or absolute (e.g. this will increase your risk by 1 in 10, 000)
-> a lot of research has shown that people's preference can be strongly influenced by how risk change information is presented
What is loss aversion?
What does loss aversion mean in terms of framing effects?
- Higher grades + faster learning
- Never study anything twice
- 100% sure, 100% understanding
What did Edwards et al's. (2001) study?
Researchers restricted their analysis/review to studies where the choices that people were being asked to make had a clear personal relevance to the participant i.e. they could potentially be faced with that choice in real life
Results = Sig. framing effect in only one of four studies conducted on personally relevant health choices. This significant effect was found in McNeil et al. (1982)
--> overall the study shows that it is not easy to replicate this framing effect
What did Moxey et al. (2003) study?
•Across 5 expts (including McNeil et al., 1982) Ps were sig more likely to choose surgery over less risky treatments in positive frame (RR=1.51,sig)
•Framing effects less obvious for immunisation (RR=1.06, ns) and medication treatment decision (RR=1.70, ns) scenarios
•Effects less convincing in studies of good methodological quality (as they control for more variables) and/or examining actual decisions
What is the summary of positive vs. negative framing?
•Risky options (like surgery) less attractive than safer alternatives (like radiation) when failure rates highlighted
•No definitive evidence that protective treatments (like vaccines/ screening tests) rated more attractive when success rates highlighted (small/ heterogeneous effect sizes)
•Hypothetical lab studies may lack ecological validity and over-estimate the effect size (very few studies have tested framing effects for decisions with real consequences)
- Real world studies = effect sizes much smaller; harder to do/ get research done; smaller samples producer smaller effects -> this means it’s harder to conclude that there is no effect
Regarding relative vs absolute framing formats, which format do people generally perceive more favourably?
= fairly robust finding
What are 2 possible issues with relative risk framing?
(Malenka et al., 1993)
= Ps may be mislead by the info because of their statistical innumeracy
2. Relative risk framing can lead to sub-optimal decision making (Baron et al., 1988, Covey, 2011)
What did Covey (2007) study?
= strong evidence that current research shows that participants favour treatments where benefits described in relative terms
--> This type of effect/bias was also found in doctors and health professionals as well - not limited to the general population (shows that even those who are in the medical field are susceptible to this effect)
What did Baron et al. (1998) study?
--> perhaps omission bias plays a role here
--> also may be a proportionality bias 50% is higher than 25%
What is the summary regarding relative risks vs. absolute risks?
• Risk changes have more impact (tend to give treatments more favourable evaluations) if presented in relative rather than absolute terms
• Proportion dominance can lead people to make inferior choices about their health care
What did Gurm and Litaker (2000) study?
Method: Presented patients with one of two videos describing angioplasty and its associated risks
Results: patients were more likely to opt for treatments when the video framed the procedure as 99% safe, compared with there being a likelihood of complication of 1 in 100
(own reading)
What did Armstrong et al. (2002) find?
Method: presented risk information using survival or mortality terms
Results: people who received survival terms were significantly more accurate in answering questions about the information, as well as being significantly more likely to opt for the treatment in question, than people who received mortality terms
(own reading)
What did Peters et al. (2010) find? What is a limitation with this study?
However, the generalizability of the findings is limited due to the use of non-patients presented with a hypothetical scenario in which they were told to imagine they had a headache. Given the design, it is unclear whether observed differences would translate into clinically significant differences in patient behaviours.
What did O'Connor et al. (1996) find?
What did Zipkin (2014) highlight?
In contrast, Zipkin (2014) found that people more accurately perceived risk differences when presented with absolute compared to relative risks. This suggests that health care risks should be presented in absolute format to lessen the framing effect bias and make better health decisions.
The question on the page originate from the summary of the following study material:
- A unique study and practice tool
- Never study anything twice again
- Get the grades you hope for
- 100% sure, 100% understanding