- see dave notes [email](https://mail.google.com/mail/u/0/#inbox/FMfcgzGmvTpRTpBjFFHnSXNjcDQfTHQf)
## Raw data summary
676605 obs 16 cols
| | NUnique| PercentMissing| Mean| SD| Min| Median| Max| Histogram|
|:------------|-------:|--------------:|--------:|-------:|----:|--------:|--------:|----------:|
|id | 34286| 0| 17256.59| 9962.41| 1| 17219.00| 34541| ▇▇▇▇▇▇▇▇▇▇|
|item_num | 45| 0| 26.74| 13.17| 1| 31.00| 45| ▄▃▄▃▃▃▄▇▆▇|
|country | 16| 0| 8.45| 4.60| 1| 8.00| 16| ▆▇▃▇▃▇▃▆▃▇|
|rt | 53066| 0| 12.01| 95.91| 0.00| 7.33| 52703.13| ▇|
|accimp | 6| 3| 4.17| 1.00| 1| 4.00| 5| ▁▂▅▇|
|conservative | 11| 4| 5.22| 2.98| 1| 5.00| 10| ▇▃▄▃▆▄▃▄▂▅|
|age | 82| 0| 40.23| 16.04| 18| 38.00| 99| ▇▅▅▄▃▂▂|
|gender | 2| 0| 1.49| 0.50| 1| 1.00| 2| ▇▇|
|condition | 4| 0| 2.50| 1.12| 1| 2.00| 4| ▇▇▇▇|
|crt_acc | 5| 3| 0.62| 0.92| 0| 0.00| 3| ▇▂▁▁|
|rating | 6| 0| 0.45| 0.36| 0.00| 0.40| 1.00| ▇▃▃▄▄▄|
|female | 2| 0| 0.49| 0.50| 0| 0.00| 1| ▇▇|
|college | 3| 4| 0.50| 0.50| 0| 1.00| 1| ▇▇|
|real | 2| 0| 0.50| 0.50| 0| 1.00| 1| ▇▇|
|perc_acc | 707| 0| 0.44| 0.17| 0.10| 0.43| 0.84| ▂▄▇▆▅▅▆▆▂▁|
- 16 countries (treat each country as one "study")
- 34286 subjects
- all subject ids are unique (no overlaps across studies)
- 4 conditions (each country has all four conditions)
- condition 1: control/sharing baseline
- condition 2: sharing with accuracy nudge
- condition 3: accuracy judgments
- condition 4: sharing with tips
- Dave: so for looking at corr between CRT and truth discernment, you'll want to just look at condition 3 (and if you wanted to replicate the analysis for our under-review paper, youd compare 1 and 2)
- trial range: 1 to 20
- `crt_acc`: 0, 1, 2, 3
- `rt`: seconds
- `rating`: choice (`[0.0 0.2 0.6 0.8 0.4 1.0]`): to dichotomize around midpoint 0.5
- `perc_acc`: accuracy rating for headline
- `accimp`: subject's rating for how important accuracy is
- `condition`: condition (1 to 4)
| | NUnique| PercentMissing| Mean| SD| Min| Median| Max| Histogram|
|:---------|-------:|--------------:|--------:|-------:|---:|--------:|-----:|----------:|
|country | 16| 0| 8.45| 4.60| 1| 8.00| 16| ▆▇▃▇▃▇▃▆▃▇|
|condition | 4| 0| 2.50| 1.12| 1| 2.00| 4| ▇▇▇▇|
|id | 34286| 0| 17261.09| 9961.69| 1| 17251.50| 34541| ▇▇▇▇▇▇▇▇▇▇|
|N | 20| 0| 19.73| 1.91| 1| 20.00| 20| ▇|
```r
# subjects per condition
> dt1[, n_distinct(id), keyby = .(condition)]
condition V1
1: 1 8631
2: 2 8548
3: 3 8527
4: 4 8580
# subjects per country
country V1
1: 1 2099
2: 2 2115
3: 3 2215
4: 4 2348
5: 5 2124
6: 6 2089
7: 7 2172
8: 8 2105
9: 9 2086
10: 10 2279
11: 11 2257
12: 12 2072
13: 13 1949
14: 14 2108
15: 15 2122
16: 16 2146
# trials
> dt1[, .N, keyby = .(country, condition, id)][, summary(N)]
Min. 1st Qu. Median Mean 3rd Qu. Max.
1.00 20.00 20.00 19.73 20.00 20.00
# subjects per conditon per country
> dt1[, n_distinct(id), keyby = .(condition, country)]
condition country V1
1: 1 1 535
2: 1 2 505
3: 1 3 562
4: 1 4 595
5: 1 5 538
6: 1 6 522
7: 1 7 534
8: 1 8 536
9: 1 9 487
10: 1 10 588
11: 1 11 600
12: 1 12 524
13: 1 13 495
14: 1 14 534
15: 1 15 541
16: 1 16 535
17: 2 1 518
18: 2 2 532
19: 2 3 539
20: 2 4 571
21: 2 5 550
22: 2 6 522
23: 2 7 571
24: 2 8 537
25: 2 9 553
26: 2 10 560
27: 2 11 544
28: 2 12 504
29: 2 13 487
30: 2 14 513
31: 2 15 517
32: 2 16 530
33: 3 1 509
34: 3 2 527
35: 3 3 545
36: 3 4 604
37: 3 5 516
38: 3 6 502
39: 3 7 538
40: 3 8 512
41: 3 9 536
42: 3 10 556
43: 3 11 567
44: 3 12 526
45: 3 13 480
46: 3 14 521
47: 3 15 540
48: 3 16 548
49: 4 1 537
50: 4 2 551
51: 4 3 569
52: 4 4 578
53: 4 5 520
54: 4 6 543
55: 4 7 529
56: 4 8 520
57: 4 9 510
58: 4 10 575
59: 4 11 546
60: 4 12 518
61: 4 13 487
62: 4 14 540
63: 4 15 524
64: 4 16 533
condition country V1
```