## candidate preference
```r
> d0[, summ(feols(diff_lean_bidentrump2 ~ scale(pfc_score_mean) * scale(topicC), data = .SD, se = "HC1")), keyby = .(condition, lean)]
condition lean term result
<char> <char> <char> <char>
1: proHarris Harris supporter (Intercept) b = 0.56 [-0.26, 1.39], p = .177
2: proHarris Harris supporter scale(pfc_score_mean) b = -0.20 [-0.78, 0.37], p = .482
3: proHarris Harris supporter scale(topicC) b = 0.49 [-0.34, 1.31], p = .247
4: proHarris Harris supporter scale(pfc_score_mean) × scale(topicC) b = 0.21 [-0.36, 0.78], p = .473
5: proHarris Trump supporter (Intercept) b = 2.93 [1.73, 4.12], p < .001
6: proHarris Trump supporter scale(pfc_score_mean) b = -0.69 [-3.25, 1.87], p = .597
7: proHarris Trump supporter scale(topicC) b = 1.09 [-0.11, 2.29], p = .074
8: proHarris Trump supporter scale(pfc_score_mean) × scale(topicC) b = -0.29 [-2.87, 2.29], p = .827
9: proTrump Harris supporter (Intercept) b = 2.03 [1.24, 2.81], p < .001
10: proTrump Harris supporter scale(pfc_score_mean) b = -0.38 [-1.19, 0.44], p = .367
11: proTrump Harris supporter scale(topicC) b = 0.32 [-0.46, 1.10], p = .421
12: proTrump Harris supporter scale(pfc_score_mean) × scale(topicC) b = 0.16 [-0.65, 0.98], p = .697
13: proTrump Trump supporter (Intercept) b = 0.54 [-0.18, 1.26], p = .141
14: proTrump Trump supporter scale(pfc_score_mean) b = 0.75 [0.00, 1.49], p = .050 # barely
15: proTrump Trump supporter scale(topicC) b = -0.10 [-0.82, 0.62], p = .791
16: proTrump Trump supporter scale(pfc_score_mean) × scale(topicC) b = -0.26 [-1.01, 0.49], p = .495
```
## vote likelihood
```r
> d0[, summ(feols(diff_vote_chance ~ scale(pfc_score_mean) * scale(topicC), data = .SD, se = "HC1")), keyby = .(condition, lean)]
Key: <condition, lean>
condition lean term result
<char> <char> <char> <char>
1: proHarris Harris supporter (Intercept) b = 2.46 [1.71, 3.21], p < .001
2: proHarris Harris supporter scale(pfc_score_mean) b = -0.50 [-1.08, 0.08], p = .093
3: proHarris Harris supporter scale(topicC) b = 0.09 [-0.66, 0.84], p = .808
4: proHarris Harris supporter scale(pfc_score_mean) × scale(topicC) b = -0.76 [-1.33, -0.18], p = .010 # ??
5: proHarris Trump supporter (Intercept) b = 0.28 [-0.72, 1.28], p = .579
6: proHarris Trump supporter scale(pfc_score_mean) b = 0.12 [-0.82, 1.06], p = .807
7: proHarris Trump supporter scale(topicC) b = -0.52 [-1.52, 0.47], p = .304
8: proHarris Trump supporter scale(pfc_score_mean) × scale(topicC) b = 0.28 [-0.67, 1.22], p = .563
9: proTrump Harris supporter (Intercept) b = 0.56 [-0.28, 1.40], p = .191
10: proTrump Harris supporter scale(pfc_score_mean) b = -0.41 [-1.22, 0.39], p = .313
11: proTrump Harris supporter scale(topicC) b = 0.07 [-0.78, 0.92], p = .871
12: proTrump Harris supporter scale(pfc_score_mean) × scale(topicC) b = -1.01 [-1.82, -0.20], p = .015 # ??
13: proTrump Trump supporter (Intercept) b = 2.47 [1.52, 3.43], p < .001
14: proTrump Trump supporter scale(pfc_score_mean) b = 0.00 [-0.87, 0.87], p = .997
15: proTrump Trump supporter scale(topicC) b = -0.21 [-1.17, 0.74], p = .661
16: proTrump Trump supporter scale(pfc_score_mean) × scale(topicC) b = 0.42 [-0.46, 1.29], p = .350
```
## vote choice
vote harris
```r
> d0[, summ(feols(diff_vote_harris ~ scale(pfc_score_mean) * scale(topicC), data = .SD, se = "HC1")), keyby = .(condition, lean)]
condition lean term result
<char> <char> <char> <char>
1: proHarris Harris supporter (Intercept) b = 2.37 [1.07, 3.66], p < .001
2: proHarris Harris supporter scale(pfc_score_mean) b = -0.10 [-1.12, 0.92], p = .850
3: proHarris Harris supporter scale(topicC) b = 1.28 [-0.01, 2.57], p = .052
4: proHarris Harris supporter scale(pfc_score_mean) × scale(topicC) b = -0.63 [-1.64, 0.38], p = .222
5: proHarris Trump supporter (Intercept) b = 1.98 [0.59, 3.38], p = .005
6: proHarris Trump supporter scale(pfc_score_mean) b = 0.79 [-0.38, 1.96], p = .183
7: proHarris Trump supporter scale(topicC) b = -0.62 [-2.02, 0.78], p = .383
8: proHarris Trump supporter scale(pfc_score_mean) × scale(topicC) b = 1.10 [-0.07, 2.27], p = .065 # ??
9: proTrump Harris supporter (Intercept) b = -2.19 [-3.74, -0.64], p = .006
10: proTrump Harris supporter scale(pfc_score_mean) b = 0.02 [-1.49, 1.54], p = .975
11: proTrump Harris supporter scale(topicC) b = 0.23 [-1.31, 1.76], p = .772
12: proTrump Harris supporter scale(pfc_score_mean) × scale(topicC) b = 0.62 [-0.92, 2.15], p = .428
13: proTrump Trump supporter (Intercept) b = 0.79 [-0.23, 1.81], p = .129
14: proTrump Trump supporter scale(pfc_score_mean) b = -0.77 [-2.01, 0.47], p = .221
15: proTrump Trump supporter scale(topicC) b = -0.11 [-1.13, 0.92], p = .840
16: proTrump Trump supporter scale(pfc_score_mean) × scale(topicC) b = -1.07 [-2.31, 0.17], p = .091 # ??
```
vote trump
```r
> d0[, summ(feols(diff_vote_trump ~ scale(pfc_score_mean) * scale(topicC), data = .SD, se = "HC1")), keyby = .(condition, lean)]
condition lean term result
<char> <char> <char> <char>
1: proHarris Harris supporter (Intercept) b = -0.36 [-1.05, 0.33], p = .304
2: proHarris Harris supporter scale(pfc_score_mean) b = 0.21 [-0.25, 0.67], p = .360
3: proHarris Harris supporter scale(topicC) b = -0.03 [-0.72, 0.66], p = .924
4: proHarris Harris supporter scale(pfc_score_mean) × scale(topicC) b = 0.22 [-0.24, 0.67], p = .350
5: proHarris Trump supporter (Intercept) b = -1.57 [-3.39, 0.26], p = .093
6: proHarris Trump supporter scale(pfc_score_mean) b = -0.68 [-2.04, 0.69], p = .331
7: proHarris Trump supporter scale(topicC) b = 0.61 [-1.21, 2.44], p = .509
8: proHarris Trump supporter scale(pfc_score_mean) × scale(topicC) b = -1.58 [-2.95, -0.21], p = .023
9: proTrump Harris supporter (Intercept) b = 1.22 [0.32, 2.12], p = .008
10: proTrump Harris supporter scale(pfc_score_mean) b = -0.88 [-1.73, -0.02], p = .045 # barely
11: proTrump Harris supporter scale(topicC) b = 0.56 [-0.33, 1.45], p = .216
12: proTrump Harris supporter scale(pfc_score_mean) × scale(topicC) b = -0.61 [-1.45, 0.22], p = .149
13: proTrump Trump supporter (Intercept) b = 0.89 [-0.67, 2.45], p = .265
14: proTrump Trump supporter scale(pfc_score_mean) b = 1.31 [-0.36, 2.97], p = .125
15: proTrump Trump supporter scale(topicC) b = 0.15 [-1.41, 1.71], p = .850
16: proTrump Trump supporter scale(pfc_score_mean) × scale(topicC) b = 1.48 [-0.19, 3.15], p = .083 # ??
```