# Observed RTs ~ Likert sharing intentions ```r # quadratic effect of ratingSquare on rt > m1a <- lmer(rt ~ ratingC + ratingSquareC + (1 + ratingC + ratingSquareC | study / id), dt1) > summaryh(m1a) term results 1: (Intercept) b = 8.27, SE = 0.44, t(5) = 18.93, p < .001, r = 0.99 2: ratingC b = −0.18, SE = 0.11, t(5) = −1.61, p = .165, r = 0.57 3: ratingSquareC b = −4.37, SE = 0.32, t(4) = −13.80, p < .001, r = 0.99 ``` ![[Pasted image 20220131231642.png]] ![[Pasted image 20220131232747.png]] # DDM model-predicted RTs ~ Likert sharing intentions ```r # quadratic effect of ratingSquare on DDM predicted rt m3 <- lmer(rt_pred ~ ratingC + ratingSquareC + (1 + ratingC + ratingSquareC | study / id), dt3) > summaryh(m3) term results 1: (Intercept) b = 7.22, SE = 0.41, t(5) = 17.44, p < .001, r = 0.99 2: ratingC b = 0.09, SE = 0.03, t(5) = 2.76, p = .041, r = 0.78 3: ratingSquareC b = −0.42, SE = 0.06, t(5) = −6.77, p = .001, r = 0.95 ``` ![[Pasted image 20220131222152.png]] # Observed RT ~ model-predicted RT for different Likert ratings and studies - fanning out as RT increases (> 10 s) - fewer longer RTs to fit model to, hence worse model fit ![[Pasted image 20220131222805.png]] ![[Pasted image 20220131233950.png]] ```r # correlations > dt3_avg[, .(r = cor(rt, rt_pred)), keyby = .(rating)] rating r 1: 0.0 0.9343968 2: 0.2 0.8825916 3: 0.4 0.8758244 4: 0.6 0.8857065 5: 0.8 0.8709037 6: 1.0 0.8761314 > dt3_avg[, .(r = cor(rt, rt_pred)), keyby = .(study, rating)] study rating r 1: B 0.0 0.9552928 2: B 0.2 0.8704365 3: B 0.4 0.8542955 4: B 0.6 0.8545920 5: B 0.8 0.8510144 6: B 1.0 0.8189048 7: C 0.0 0.9300002 8: C 0.2 0.8985862 9: C 0.4 0.8689930 10: C 0.6 0.8810205 11: C 0.8 0.8437642 12: C 1.0 0.8367263 13: G 0.0 0.9063222 14: G 0.2 0.8645948 15: G 0.4 0.8530465 16: G 0.6 0.8683215 17: G 0.8 0.8548013 18: G 1.0 0.8583102 19: H 0.0 0.9147807 20: H 0.2 0.8676737 21: H 0.4 0.8819867 22: H 0.6 0.8809585 23: H 0.8 0.8761860 24: H 1.0 0.8355558 25: L 0.0 0.9751849 26: L 1.0 0.9577894 27: P 0.0 0.8928729 28: P 0.2 0.8973224 29: P 0.4 0.9025524 30: P 0.6 0.9124544 31: P 0.8 0.8862142 32: P 1.0 0.8517898 study rating r ``` # One study (L) has binary sharing intentions ```r # mean RT in each study platform study condition rt choice 1: Lucid G 0 8.491321 0.3847112 2: Lucid G 1 8.798396 0.3522638 3: Lucid L 0 6.979766 0.4609744 # binary 4: Lucid L 1 7.095794 0.4575841 # binary 5: Lucid P 0 5.716850 0.4358538 6: Lucid P 1 5.984047 0.4331498 7: MTurk B 0 8.377431 0.2301779 8: MTurk B 1 8.428912 0.2358963 9: MTurk C 0 7.885660 0.2706938 10: MTurk C 1 8.481407 0.2571657 11: MTurk H 0 6.920213 0.4085294 12: MTurk H 1 6.960239 0.3968796 binary condition rt rating2 1: -0.5 0 7.792818 0.3266990 2: -0.5 1 7.548500 0.3453438 3: 0.5 0 6.979766 0.4609744 # binary 4: 0.5 1 7.095794 0.4575841 # binary # driftb1 ~ condition > d[, summaryh(lm(driftb1 ~ conditionEC)), keyby = .(binarized, Platform, study)][order(term)] binarized Platform study term results 1: -0.5 Lucid G (Intercept) b = 0.05, SE = 0.03, t(710) = 1.98, p = .048, r = 0.07 2: -0.5 Lucid P (Intercept) b = 0.38, SE = 0.08, t(386) = 4.62, p < .001, r = 0.23 3: -0.5 MTurk B (Intercept) b = 0.12, SE = 0.03, t(918) = 4.08, p < .001, r = 0.13 4: -0.5 MTurk C (Intercept) b = 0.17, SE = 0.03, t(981) = 6.64, p < .001, r = 0.21 5: -0.5 MTurk H (Intercept) b = 0.48, SE = 0.03, t(1741) = 15.83, p < .001, r = 0.36 6: 0.5 Lucid L (Intercept) b = 0.49, SE = 0.06, t(590) = 8.43, p < .001, r = 0.33 # binary 7: -0.5 Lucid G conditionEC b = 0.16, SE = 0.05, t(710) = 2.86, p = .004, r = 0.11 8: -0.5 Lucid P conditionEC b = 0.14, SE = 0.16, t(386) = 0.85, p = .394, r = 0.04 9: -0.5 MTurk B conditionEC b = 0.33, SE = 0.06, t(918) = 5.44, p < .001, r = 0.18 10: -0.5 MTurk C conditionEC b = 0.27, SE = 0.05, t(981) = 5.12, p < .001, r = 0.16 11: -0.5 MTurk H conditionEC b = 0.03, SE = 0.06, t(1741) = 0.42, p = .673, r = 0.01 12: 0.5 Lucid L conditionEC b = 0.25, SE = 0.12, t(590) = 2.18, p = .029, r = 0.09 # binary # boundary ~ condition > d[, summaryh(lm(B ~ conditionEC)), keyby = .(binarized, Platform, study)][order(term)] binarized Platform study term results 1: -0.5 Lucid G (Intercept) b = 3.42, SE = 0.12, t(710) = 29.60, p < .001, r = 0.74 2: -0.5 Lucid P (Intercept) b = 2.62, SE = 0.07, t(386) = 37.57, p < .001, r = 0.89 3: -0.5 MTurk B (Intercept) b = 3.45, SE = 0.08, t(918) = 43.66, p < .001, r = 0.82 4: -0.5 MTurk C (Intercept) b = 3.24, SE = 0.08, t(981) = 41.75, p < .001, r = 0.80 5: -0.5 MTurk H (Intercept) b = 2.64, SE = 0.06, t(1741) = 46.51, p < .001, r = 0.74 6: 0.5 Lucid L (Intercept) b = 2.89, SE = 0.10, t(590) = 30.14, p < .001, r = 0.78 # binary (mean boundary similar to other studies) 7: -0.5 Lucid G conditionEC b = 0.08, SE = 0.23, t(710) = 0.33, p = .744, r = 0.01 8: -0.5 Lucid P conditionEC b = 0.04, SE = 0.14, t(386) = 0.25, p = .801, r = 0.01 9: -0.5 MTurk B conditionEC b = −0.02, SE = 0.16, t(918) = −0.13, p = .896, r = 0 10: -0.5 MTurk C conditionEC b = −0.10, SE = 0.16, t(981) = −0.67, p = .504, r = −0.02 11: -0.5 MTurk H conditionEC b = −0.13, SE = 0.11, t(1741) = −1.16, p = .247, r = −0.03 12: 0.5 Lucid L conditionEC b = −0.04, SE = 0.19, t(590) = −0.20, p = .838, r = −0.01 # binary ```