Normalised MEC (fish example)
 Share
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

View only
 
 
ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
Notes
2
I think I've done this correctly, but I'm not sure. I tried to follow the formula here: https://www.cs.cmu.edu/~epxing/Class/10810-07/lectures/normLec.pdf (slides 9 and 11). I didn't know for sure what should count as the population variance and population mean, so I used the variances and means across all theories and options. It's possible the "standardisation" approaches here are relevant, but I think that's something separate: https://medium.com/@swethalakshmanan14/how-when-and-why-should-you-normalize-standardize-rescale-your-data-3f083def38ff
3
Reasons to think I did it correctly: The variances of the normalised functions come out as equal, which is the goal. The pattern of results in the three-option case seems to make sense; playing around with the T1 credence (and by extension with the T2 credence) suggested this basically gives an answer that's less favourable to T1 than the non-normalised answer was, but can still go T1's way if T1 credence is high enough, and especially can mandate the compromise egg curry; this seems to make sense, because normalisation reduces the disproportionate apparent "stakes" for T1.
4
Things I'm still confused about: Does it make sense that in the two-option case, post-normalisation, it seems that all that matters is the credences (it basically seems to be My Favourite Theory), based on playing around with the credences? Does it matter that the normalised CW scores are negative in most cases? (I don't think so, as utility functions can be given a positive affine transformation.) Can/should I fix that? Why are the ranges the same (but opposite sign) in the two option case, but different (though similar) in the one option case?
5
6
With three options
7
Credence assigned
Non-normalised choice-worthiness
Expected CW of option
Mean of theory's (T's) CW function
Variance of T's CW function
Overall mean
Overall variance
Normalised CW
Expected CW of option
Variance of T's normalised CW functions (just to check I did my calculations right)
Ranges of normalised CW
8
Fish curryT10.3205-90-22.05-28.333333332858.333333-10.333333331534.666667-52.75736418-101.5282061534.66666769.61041107
9
T20.6795107.6666666676.333333333-124.53196591534.66666777.83247868
10
Tofu curryT10.32055516.85304689-132.1052385
11
T20.67955-202.3644446
12
Egg curryT10.320505.43613.18934104-101.5471547
13
T20.67958-155.6649574
14
15
Two options only
16
Credence assigned
Non-normalised choice-worthiness
Expected CW of option
Mean of theory's (T's) CW function
Variance of T's CW function
Overall mean
Overall variance
Normalised CW
Expected CW of option
Variance of T's normalised CW functions (just to check I did my calculations right)
Ranges of normalised CW
17
Fish curryT10.25-90-15-42.54512.5-17.52341.666667-46.82386793-165.68445572341.666667-68.43488389
18
T20.75107.512.5-205.30465172341.66666768.43488389
19
Tofu curryT10.255521.61101597-199.9018977
20
T20.755-273.7395356
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Loading...