Academia.eduAcademia.edu

Figure 5 – uploaded by Rebecca McDonald

See full PDFdownloadDownload figure
Figure 3a. Randomly selected participant from the bottom 5% consistency score.

Figure 3 a. Randomly selected participant from the bottom 5% consistency score.

Related Figures (12)

To measure consistency of each individual’s ranking across different conditions we adopted Krippendorff's coefficient (Krippendorff 2004).
To measure consistency of each individual’s ranking across different conditions we adopted Krippendorff's coefficient (Krippendorff 2004).
Figure 1. Average ranking and corresponding 95% Confidence Intervals for all conditions. density distributions of ranks for each data type and each condition. ind corresponding 95% Confidence Intervals, and Figure 2a, Figure 2b, and Figure 2c for
Figure 1. Average ranking and corresponding 95% Confidence Intervals for all conditions. density distributions of ranks for each data type and each condition. ind corresponding 95% Confidence Intervals, and Figure 2a, Figure 2b, and Figure 2c for
Figure 2a. Density distribution of ranks for Tier 1 data: Banking and Medical Records Data.
Figure 2a. Density distribution of ranks for Tier 1 data: Banking and Medical Records Data.
‘igure 2b. Density distribution of ranks for Tier 2 data: Browsing History, Mobile Phone GPS
‘igure 2b. Density distribution of ranks for Tier 2 data: Browsing History, Mobile Phone GPS
igure 2c. Density distribution of ranks for Tier 3 data: Electricity Use, Loyalty Cards anc
igure 2c. Density distribution of ranks for Tier 3 data: Electricity Use, Loyalty Cards anc
Table 2. Break down of Krippendorff alpha consistency scores by poor, slight, fair, moderate,
Table 2. Break down of Krippendorff alpha consistency scores by poor, slight, fair, moderate,
brackets, as well as random effects, Log Likelihood. Values highlighted in bold a significant
brackets, as well as random effects, Log Likelihood. Values highlighted in bold a significant
Figure 4 demonstrates predicted effects for each data type, conditioned on all other
Figure 4 demonstrates predicted effects for each data type, conditioned on all other
(5.1%, WTP Choice; 19.5% WTP Slider). Individual consistency between ranking orders predicted higher likelinood of being willing to
(5.1%, WTP Choice; 19.5% WTP Slider). Individual consistency between ranking orders predicted higher likelinood of being willing to
below. Figure 5. Predicted amount individuals are willing to pay for different data types, out of those results of the Probit analysis. We discuss the OLS results and differences from the Probit zero. The results of the OLS regression differed from (but did not seriously contradict) the
below. Figure 5. Predicted amount individuals are willing to pay for different data types, out of those results of the Probit analysis. We discuss the OLS results and differences from the Probit zero. The results of the OLS regression differed from (but did not seriously contradict) the
Related topics:
BusinessPsychologyInternet privacy
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved