 |
|
Was this paper discussed?...
Posted on: December 18, 2021 at 22:19:35 CT
JeffB MU
Posts:
72858
Member For:
21.61 yrs
Level:
User
M.O.B. Votes:
0
The Science of Visual Data Communication: What Works
Steven L. Franconeri, Lace M. Padilla, Priti Shah, Jeffrey M. Zacks, Jessica Hullman
Jessica Hullman
https://journals.sagepub.com/stoken/default+domain/10.1177%2F15291006211051956-FREE/full
Convey risks with absolute—not relative—rates
A rational decision maker should not make different treatment decisions based on the way that medical risk is conveyed. Unfortunately, the presentation of medical-risk information has a profound influence on judgments, even for highly educated people. In a study of 235 practicing physicians who were asked to indicate how likely they were to treat patients for hypertension and hypercholesterolemia on the basis of information about the treatments; 41.3% were more likely to recommend a treatment when its effects were framed as a relative reduction in risk compared with an absolute reduction in risk (Forrow et al., 1992). Absolute risk is the overall probability of experiencing something, such as getting a disease, during some period of time. Relative risk is a comparison between the absolute risk of two groups (Natter & Berry, 2005). As Fagerlin et al. (2011) illustrated, the change in breast cancer risk from 4% to 2% associated with the drug Tamoxifen can be communicated in a relative format (a reduction of 50%) or an absolute format (an absolute reduction of 2%).
Absolute-risk formats are preferable: Many studies have found that the same treatment outcomes are perceived more favorably when communicated in terms of relative risk as opposed to absolute risk (for a meta-analysis, see Covey, 2007). Further, relative-risk framing can make changes in risk appear larger than absolute-risk framing (e.g., a 50% relative change incorrectly seems larger than a 2% absolute change; Akl et al., 2011; Baron, 1997; Forrow et al., 1992; Malenka et al., 1993). This framing bias makes physicians more likely to prescribe interventions when information about their effects is communicated using a relative-risk format (e.g., Bucher et al., 1994; Lacy et al., 2001; Naylor et al., 1992). A large body of evidence suggests that the absolute-risk format leads to the least biased decisions. Although this has not been tested with visualizations, the evidence points to absolute risk as the best choice.
Steven L. Franconeri, Lace M. Padilla, Priti Shah, Jeffrey M. Zacks, Jessica Hullman
Jessica Hullman
https://journals.sagepub.com/stoken/default+domain/10.1177%2F15291006211051956-FREE/full
Convey risks with absolute—not relative—rates
A rational decision maker should not make different treatment decisions based on the way that medical risk is conveyed. Unfortunately, the presentation of medical-risk information has a profound influence on judgments, even for highly educated people. In a study of 235 practicing physicians who were asked to indicate how likely they were to treat patients for hypertension and hypercholesterolemia on the basis of information about the treatments; 41.3% were more likely to recommend a treatment when its effects were framed as a relative reduction in risk compared with an absolute reduction in risk (Forrow et al., 1992). Absolute risk is the overall probability of experiencing something, such as getting a disease, during some period of time. Relative risk is a comparison between the absolute risk of two groups (Natter & Berry, 2005). As Fagerlin et al. (2011) illustrated, the change in breast cancer risk from 4% to 2% associated with the drug Tamoxifen can be communicated in a relative format (a reduction of 50%) or an absolute format (an absolute reduction of 2%).
Absolute-risk formats are preferable: Many studies have found that the same treatment outcomes are perceived more favorably when communicated in terms of relative risk as opposed to absolute risk (for a meta-analysis, see Covey, 2007). Further, relative-risk framing can make changes in risk appear larger than absolute-risk framing (e.g., a 50% relative change incorrectly seems larger than a 2% absolute change; Akl et al., 2011; Baron, 1997; Forrow et al., 1992; Malenka et al., 1993). This framing bias makes physicians more likely to prescribe interventions when information about their effects is communicated using a relative-risk format (e.g., Bucher et al., 1994; Lacy et al., 2001; Naylor et al., 1992). A large body of evidence suggests that the absolute-risk format leads to the least biased decisions. Although this has not been tested with visualizations, the evidence points to absolute risk as the best choice.
Please explain why this message is being reported.
MESSAGE THREAD
Was this paper discussed?... - JeffB MU - 12/18 22:19:35