OSCR

The neural dynamics of political socio-pragmatic violations: an ERP study.

Code ↔ Paper

2 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 2 matches
  1. [1] § Materials and methods › Statistical analysis › ERP analysis ↔ polpejEEG_repo.R, lines 143–181 · score 0.69 · Bonferroni Holm, NoQM, lme4, bobyqa, optimizer, IC
  2. [2] § Materials and methods › Statistical analysis › Lexeme characteristics and SDO ↔ polpejEEG_repo.R, lines 143–181 · score 0.65 · Bonferroni Holm, NoQM, Word frequencies, syllables, position, channel

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

R · 627 lines · 22 KB · no license · 2 matches

  1. library(lme4)
  2. library(glmmTMB)
  3. library(emmeans)
  4. library(ggplot2)
  5. library(sjPlot)
  6. library(lmerTest)
  7. library(lattice)
  8. library(rlist)
  9. library(DHARMa)
  10. library(ggsignif)
  11. library(stringr)
  12. library(dplyr)
  13. library(tidyr)
  14. library(ggeffects)
  15. library(sjstats)
  16. library(ggpubr)
  17. library(effectsize)
  20. # set working directory
  21. setwd("C:/Users/maure/Desktop/PolPejEEG/Export")
  23. # read ERP data
  25. N400_df <- read.csv("./csvs/N400_df.csv")
  26. P300_df <- read.csv("./csvs/P300_df.csv")
  27. LPP_df <- read.csv("./csvs/LPP_df.csv")
  30. # set lmem and notational options
  31. emm_options(pbkrtest.limit = 3000)
  32. options(scipen=999)
  35. # bring dfs into long format and drop rows containing nans
  36. N400_df_long <- pivot_longer(N400_df, c("Cz", "Pz"), names_to = "Channel", values_to = "Amplitude")
  37. N400_df_long <- drop_na(N400_df_long, "Pej.Value")
  38. N400_df_long <- rename(N400_df_long, Amplitude_N400=Amplitude)
  40. n_df <- N400_df_long %>%
  41. filter(Channel=="Cz") %>%
  42. group_by(Participant, Coherence, Quotation) %>%
  43. count()
  44. n_df[n_df$n < 15, ]
  46. length(unique(N400_df$Participant))
  47. length(unique(N400_df_long$Participant))
  49. P300_df_long <- pivot_longer(P300_df, c("Cz", "Pz"), names_to = "Channel", values_to = "Amplitude")
  50. P300_df_long <- drop_na(P300_df_long, "Pej.Value")
  51. P300_df_long <- rename(P300_df_long, Amplitude_P300=Amplitude)
  53. LPP_df_long <- pivot_longer(LPP_df, c("P3", "Pz", "P4"), names_to = "Channel", values_to = "Amplitude")
  54. LPP_df_long <- drop_na(LPP_df_long)
  55. LPP_df_long <- rename(LPP_df_long, Amplitude_LPP=Amplitude)
  57. Amplitude_P300 <- P300_df_long$Amplitude_P300
  58. Amplitude_LPP <- LPP_df_long$Amplitude_LPP
  60. # bind dfs
  61. df_long <- cbind(N400_df_long, Amplitude_P300)
  63. # scale and center numerical variables
  64. df_long$Pej.Value <- scale(df_long$Pej.Value, center=T)
  65. df_long$Amplitude_N400_unscaled <- df_long$Amplitude_N400
  66. df_long$Amplitude_N400 <- scale(df_long$Amplitude_N400, center=T)
  67. df_long$Amplitude_P300_unscaled <- df_long$Amplitude_P300
  68. df_long$Amplitude_P300 <- scale(df_long$Amplitude_P300, center=T)
  69. df_long$SDO_mean <- scale(df_long$SDO_mean, center=T)
  71. LPP_df_long$Amplitude_LPP_unscaled <- LPP_df_long$Amplitude_LPP
  72. LPP_df_long$Amplitude_LPP <- scale(LPP_df_long$Amplitude_LPP, center=T)
  73. LPP_df_long$Pej.Value <- scale(LPP_df_long$Pej.Value, center=T)
  75. # factorize categorical variables
  76. df_long$Coherence <- as.factor(df_long$Coherence)
  77. df_long$Quotation <- as.factor(df_long$Quotation)
  79. LPP_df_long$Coherence <- as.factor(LPP_df_long$Coherence)
  80. LPP_df_long$Quotation <- as.factor(LPP_df_long$Quotation)
  84. # perform effect coding
  85. contrasts(df_long$Coherence) <- "contr.sum"
  86. contrasts(df_long$Quotation) <- "contr.sum"
  88. contrasts(LPP_df_long$Coherence) <- "contr.sum"
  89. contrasts(LPP_df_long$Quotation) <- "contr.sum"
  92. ## extract and save raw amplitude parameters
  93. m_sd_N400 <- df_long %>% drop_na() %>% group_by(Coherence, Quotation, Channel) %>%
  94. summarize(mean_amplitude = mean
  95. (Amplitude_N400_unscaled), sd_amplitude = sd(Amplitude_N400_unscaled))
  97. m_sd_P300 <- df_long %>% drop_na() %>% group_by(Coherence, Quotation, Channel) %>%
  98. summarize(mean_amplitude = mean
  99. (Amplitude_P300_unscaled), sd_amplitude = sd(Amplitude_P300_unscaled))
  101. m_sd_LPP <- LPP_df_long %>% drop_na() %>% group_by(Coherence, Quotation, Channel) %>%
  102. summarize(mean_amplitude = mean
  103. (Amplitude_LPP_unscaled), sd_amplitude = sd(Amplitude_LPP_unscaled))
  105. m_sd_df <- as.data.frame(rbind(m_sd_N400, m_sd_P300, m_sd_LPP))
  106. write.csv(m_sd_df, file="./csvs/m_sd_df.csv")
  110. df_long_1ch <- df_long %>% filter(Channel == "Cz")
  111. word_df <- read.csv("./word_data/word_frequencies_dwds_gegenwartskorpus.csv")
  112. word_df_clean <- word_df[word_df$AmbigUsage!="Y",]
  113. df_word_freq <- merge(df_long_1ch, word_df[, -3], by="Word")
  114. contrasts(df_word_freq$Coherence) <- "contr.treatment"
  115. df_word_freq_clean <- df_word_freq %>% filter(AmbigUsage!="Y")
  119. df_word_freq_clean$logWordsFrequencyPerMillion <- log(df_word_freq_clean$WordsFrequencyPerMillion+1)
  120. df_word_freq_clean %>% group_by(Coherence) %>%
  121. summarize(M = mean(WordsFrequencyPerMillion),
  122. SD = sd(WordsFrequencyPerMillion),
  123. Median = median(WordsFrequencyPerMillion),
  124. logM = mean(logWordsFrequencyPerMillion),
  125. logSD = sd(logWordsFrequencyPerMillion),
  126. logMedian = median(logWordsFrequencyPerMillion))
  130. pej_df_word <- df_long_1ch %>% group_by(Word) %>%
  131. summarize(mean_pej = mean(Pej.Value))
  133. word_df <- left_join(word_df, pej_df_word, by="Word")
  134. word_df
  135. write.csv(word_df, file="./word_data/word_characteristics_supplementary.csv")
  137. C_clean <- word_df_clean[word_df_clean$Coherence=="C",]
  138. IC_clean <- word_df_clean[word_df_clean$Coherence=="IC",]
  139. C <- word_df[word_df$Coherence=="C",]
  140. IC <- word_df[word_df$Coherence=="IC",]
  143. # t-tests of potential covariates
  145. t1 <- t.test(C_clean$WordsFrequencyPerMillion, IC_clean$WordsFrequencyPerMillion)
  146. t1
  147. sd(C_clean$WordsFrequencyPerMillion)
  148. sd(IC_clean$WordsFrequencyPerMillion)
  149. t2 <- t.test(C$n_Letters, IC$n_Letters)
  150. t2
  151. sd(C$n_Letters)
  152. sd(IC$n_Letters)
  153. t3 <- t.test(C$n_Syllables, IC$n_Syllables)
  154. t3
  155. sd(C$n_Syllables)
  156. sd(IC$n_Syllables)
  157. t4 <- t.test(C$TargetPosition, IC$TargetPosition)
  158. t4
  159. sd(C$TargetPosition)
  160. sd(IC$TargetPosition)
  161. t5 <- t.test(C$mean_pej, IC$mean_pej)
  162. t5
  163. sd(C$mean_pej)
  164. sd(IC$mean_pej)
  166. p_list <- sort(c(t1$p.value, t2$p.value, t3$p.value, t4$p.value, t5$p.value))
  168. p_list*c(5, 4, 3, 2, 1) #bonferroni-holm correction
  170. word_df_syl <- word_df %>% select(c(Word, n_Syllables))
  171. df_long <- merge(df_long, word_df_syl, by="Word")
  172. df_long_NoQM <- df_long %>% filter(Quotation=="NoQM")
  175. # build model based on n_syllables to show assess the isolated potential effect of the number of syllables
  176. syl_model <- lmer(Amplitude_N400 ~ n_Syllables + (1+n_Syllables|Participant) + (1|Word) + (1|Channel),
  177. data=df_long_NoQM, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  178. summary(syl_model)
  180. # building the models usually takes a lot of time and resources
  181. # models objects are therefore provided as .rda files and can be loaded
  184. # N400 amplitude analysis
  186. #model_full <- lmer(Amplitude_N400 ~ Coherence*Quotation*Pej.Value + (1 + Coherence*Quotation*Pej.Value|Participant) + (1 + Quotation*Pej.Value|Word) + (1 + Coherence*Quotation*Pej.Value|Channel),
  187. # data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  188. #save(model_full, file="N400_model_full.rda")
  189. #load("N400_model_full.rda")
  190. #summary(model_full)
  192. #model_nochslopes <- lmer(Amplitude_N400 ~ Coherence*Quotation*Pej.Value + (1 + Coherence*Quotation*Pej.Value||Participant) + (1 + Quotation*Pej.Value||Word) + (1|Channel),
  193. # data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  194. #save(model_nochslopes, file="N400_model_nochslopes.rda")
  195. #load("N400_model_nochslopes.rda")
  196. #summary(model_nochslopes)
  199. #model_inter_nochslopes <- lmer(Amplitude_N400 ~ Coherence*Quotation*Pej.Value + (1 + Coherence:Quotation:Pej.Value|Participant) + (1 + Quotation:Pej.Value|Word) + (1|Channel),
  200. # data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  201. #save(model_inter_nochslopes, file="N400_model_inter_nochslopes.rda")
  202. #load("N400_model_inter_nochslopes.rda")
  203. #summary(model_inter_nochslopes)
  206. #model_inter_nochslopes_nocorr <- lmer(Amplitude_N400 ~ Coherence*Quotation*Pej.Value + (1 + Coherence:Quotation:Pej.Value||Participant) + (1 + Quotation:Pej.Value||Word) + (1|Channel),
  207. # data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  208. #save(model_inter_nochslopes_nocorr, file="N400_model_inter_nochslopes_nocorr.rda")
  209. #load("N400_model_inter_nochslopes_nocorr.rda")
  210. #summary(model_inter_nochslopes_nocorr)
  214. #model <- model_inter_nochslopes_nocorr
  216. #save(model, file="N400_model.rda")
  219. model_SDO_re <- lmer(Amplitude_N400 ~ Coherence*Quotation*Pej.Value + (1 + Coherence:Quotation:Pej.Value||SDO_mean) + (1 + Quotation:Pej.Value||Word) + (1|Channel),
  220. data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  221. save(model_SDO_re, file="N400_model_SDO_re.rda")
  222. load("N400_model_SDO_re.rda")
  223. model_SDO_sum <- summary(model_SDO_re)
  224. print(model_SDO_sum)
  225. write.csv(model_SDO_sum$coefficients, "./csvs/combined_model_SDO_summary_N400.csv") # write summary to file
  229. model_SDO <- lmer(Amplitude_N400 ~ Coherence*Quotation*Pej.Value*SDO_mean + (1 + Coherence:Quotation:Pej.Value:SDO_mean||Participant) + (1 + Quotation:Pej.Value:SDO_mean||Word) + (1|Channel),
  230. data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  231. save(model_SDO, file="N400_model_SDO.rda")
  232. summary(model_SDO)
  237. ## load model
  238. load("N400_model.rda")
  240. anova(model_SDO_re, model)
  246. ## lmem summary
  248. model_sum <- summary(model)
  249. print(model_sum)
  250. write.csv(model_sum$coefficients, "./csvs/combined_model_summary_N400.csv") # write summary to file
  253. ## estimated marginal means
  254. emm_means <- emmeans(model, ~ Coherence*Quotation) # define estimated marginal means formula
  255. emm_means
  256. emm_contrasts <- contrast(emm_means, list(c(1, -1, 0, 0), c(0, 0, 1, -1), c(0, 1, 0, -1), c(1, 0, -1, 0)), adjust="holm") # define contrasts for emms
  257. emm_contrasts
  259. emm_means_df <- as.data.frame(emm_means)
  260. emm_contrasts_df <- as.data.frame(emm_contrasts) %>% mutate(contrast = str_replace_all(contrast, "C ", "C-"))
  262. #eta_squared(model, partial=TRUE)
  264. ## define plot variables
  265. means <- emm_means_df$emmean
  266. SEs <- emm_means_df$SE
  267. max_NoQM <- max((means+SEs)[1:2])
  268. max_QM <- max((means+SEs)[3:4])
  271. ## plot group differences
  272. QC_plot <- ggplot(emm_means_df, aes(x = Quotation, y = emmean, color = Coherence)) +
  273. ylim(-0.6, 0.6) +
  274. geom_point(position = position_dodge(width = 0.3), size=3) +
  275. geom_errorbar(aes(ymin = means-SEs, ymax = means+SEs),
  276. position = position_dodge(width = 0.3), # Shift error bars
  277. linewidth = 0.8, # Make error bars thicker
  278. width = 0.2) +
  279. labs(title = "",
  280. x = "Quotation",
  281. y = "N400 Amplitude (z-standardized)",
  282. color = "Coherence") +
  283. theme(text = element_text(size=14)) +
  284. geom_signif(y_position = c(max_QM+0.05, max_NoQM+0.05, max_NoQM+0.15, max_QM+0.15),
  285. xmin = c(1.925, 0.925, 0.925, 1.075),
  286. xmax = c(2.075, 1.075, 1.925, 2.075),
  287. annotation = c("*", "*", "n.s.", "***"),
  288. tip_length = 0, color="black") +
  289. scale_color_brewer(palette="Dark2") +
  290. theme(legend.position = "bottom")
  292. plot(QC_plot)
  294. ggsave("./plots/QC_plot_N400.png", dpi=300, width=1000, height=1200, units="px")
  297. ## plot three-way interaction
  298. all_plot <- plot_model(model, type="pred", terms=c("Pej.Value", "Coherence", "Quotation"), title="", axis.title=c("Pejorative Weight (z-standardized)", "N400 Amplitude (z-standardized)"), colors = "Dark2") +
  299. theme(legend.position = "bottom") + ylim(-0.6, 0.6)
  300. plot(all_plot)
  301. ggsave("./plots/all_plot_N400.png", dpi=300, width=2000, height=1200, units="px")
  304. ## plot model estimates
  305. est_plot <- plot_model(model, type="est", colors="Set1", title="", axis.lim = c(-0.25, 0.25),
  306. axis.labels = c("Coh * Quot * Pej.Weight", "Quotation * Pejorative Weight",
  307. "Coherence * Pejorative Weight", "Coherence * Quotation",
  308. "Pejorative Weight", "Quotation", "Coherence"
  309. )) + theme(axis.text.y = element_text(angle = 30)) +
  310. ylim(-0.25, 0.25)
  312. plot(est_plot)
  314. ggsave("./plots/est_plot_N400.png", dpi=300, width=1500, height=1800, units="px")
  317. ## combine plots
  318. combi_plot <- ggarrange(all_plot,
  319. ggarrange(QC_plot, est_plot, ncol=2, labels=c("B", "C")),
  320. nrow=2,
  321. labels="A")
  323. plot(combi_plot)
  325. ggsave("./plots/combi_plot_N400.png", dpi=300, width=2400, height=2400, units="px")
  328. ## Diagnostics plots
  329. sim_res <- simulateResiduals(fittedModel = model, plot = F)
  331. ### Normality of residuals
  332. plotQQunif(sim_res, testUniformity = F, testOutliers = F, testDispersion = F)
  336. ### Normality of random effects
  337. load("N400_model_inter_nochslopes.rda")
  339. diag <- plot_model(model_inter_nochslopes, type="diag")
  340. # model_inter_nochslopes is used because plot_model cannot produce
  341. # diagnostic plots for model equations contain the '||' operator
  343. plot(diag[[2]]$Participant)
  344. plot(diag[[2]]$Word)
  345. plot(diag[[2]]$Channel)
  347. ### Homoskedasticity
  348. plot(diag[[4]])
  354. ## P300
  357. ## build and save model
  359. #model_full <- lmer(Amplitude_P300 ~ Coherence*Quotation*Pej.Value + (1 + Coherence*Quotation*Pej.Value|Participant) + (1 + Quotation*Pej.Value|Word) + (1 + Coherence*Quotation*Pej.Value|Channel),
  360. # data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  361. #save(model_full, file="P300_model_full.rda")
  362. #load("P300_model_full.rda")
  363. #summary(model_full)
  365. #model_nochslopes <- lmer(Amplitude_P300 ~ Coherence*Quotation*Pej.Value + (1 + Coherence*Quotation*Pej.Value||Participant) + (1 + Quotation*Pej.Value||Word) + (1|Channel),
  366. # data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  367. #save(model_nochslopes, file="P300_model_nochslopes.rda")
  368. #load("P300_model_nochslopes.rda")
  369. #summary(model_nochslopes)
  372. #model_inter_nochslopes <- lmer(Amplitude_P300 ~ Coherence*Quotation*Pej.Value + (1 + Coherence:Quotation:Pej.Value|Participant) + (1 + Quotation:Pej.Value|Word) + (1|Channel),
  373. # data=df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  374. #save(model_inter_nochslopes, file="P300_model_inter_nochslopes.rda")
  375. #load("P300_model_inter_nochslopes.rda")
  376. #summary(model_inter_nochslopes)
  379. #model <- model_inter_nochslopes
  381. #save(model, file="P300_model.rda")
  384. ## load model
  385. load("P300_model.rda")
  387. ## model summary
  388. model_sum <- summary(model)
  389. print(model_sum)
  390. write.csv(model_sum$coefficients, "./csvs/combined_model_summary_P300.csv")
  392. ## estimated marginal means
  393. emmeans(model, pairwise ~ Quotation)
  396. emm_means <- emmeans(model, ~ Coherence*Quotation)
  397. emm_means
  398. emm_contrasts <- contrast(emm_means, list(c(1, -1, 0, 0), c(0, 0, 1, -1), c(0, 1, 0, -1), c(1, 0, -1, 0)), adjust="holm")
  399. emm_contrasts
  401. emm_means_df <- as.data.frame(emm_means)
  402. emm_contrasts_df <- as.data.frame(emm_contrasts) %>% mutate(contrast = str_replace_all(contrast, "C ", "C-"))
  404. #eta_squared(model, partial=TRUE)
  406. ## define plot variables
  407. means <- emm_means_df$emmean
  408. SEs <- emm_means_df$SE
  409. max_NoQM <- max((means+SEs)[1:2])
  410. max_QM <- max((means+SEs)[3:4])
  412. ## plot group differences
  413. QC_plot <- ggplot(emm_means_df, aes(x = Quotation, y = emmean, color = Coherence)) +
  414. ylim(-0.6, 0.6) +
  415. geom_point(position = position_dodge(width = 0.3), size=3) +
  416. geom_errorbar(aes(ymin = means-SEs, ymax = means+SEs),
  417. position = position_dodge(width = 0.3), # Shift error bars
  418. linewidth = 0.8, # Make error bars thicker
  419. width = 0.2) +
  420. labs(title = "",
  421. x = "Quotation",
  422. y = "P300 Amplitude (z-standardized)",
  423. color = "Coherence") +
  424. theme(text = element_text(size=14)) +
  425. geom_signif(y_position = c(max_QM+0.15),
  426. xmin = c(1),
  427. xmax = c(2),
  428. annotation = c("***"),
  429. tip_length = 0.1, color="black") +
  430. scale_color_brewer(palette="Dark2") +
  431. theme(legend.position = "bottom")
  433. plot(QC_plot)
  435. ggsave("./plots/QC_plot_P300.png", dpi=300, width=1000, height=1200, units="px")
  438. ## plot three-way interaction
  439. all_plot <- plot_model(model, type="pred", terms=c("Pej.Value", "Coherence", "Quotation"), title="", axis.title=c("Pejorative Weight (z-standardized)", "P300 Amplitude (z-standardized)"), colors = "Dark2") +
  440. theme(legend.position = "bottom") + ylim(-0.6, 0.6)
  441. plot(all_plot)
  442. ggsave("./plots/all_plot_P300.png", dpi=300, width=2000, height=1200, units="px")
  445. ## plot model estimates
  446. est_plot <- plot_model(model, type="est", colors="Set1", title="", axis.lim = c(-2.5, 2.5),
  447. axis.labels = c("Coh * Quot * Pej.Weight", "Quotation * Pejorative Weight",
  448. "Coherence * Pejorative Weight", "Coherence * Quotation",
  449. "Pejorative Weight", "Quotation", "Coherence"
  450. )) + theme(axis.text.y = element_text(angle = 30)) +
  451. ylim(-0.25, 0.25)
  452. plot(est_plot)
  453. ggsave("./plots/est_plot_P300.png", dpi=300, width=1500, height=1800, units="px")
  456. ## combine plots
  457. combi_plot <- ggarrange(all_plot,
  458. ggarrange(QC_plot, est_plot, ncol=2, labels=c("B", "C")),
  459. nrow=2,
  460. labels="A")
  462. plot(combi_plot)
  464. ggsave("./plots/combi_plot_P300.png", dpi=300, width=2400, height=2400, units="px")
  468. ## Diagnostics plots
  469. sim_res <- simulateResiduals(fittedModel = model, plot = F)
  471. ### Normality of residuals
  472. plotQQunif(sim_res, testUniformity = F, testOutliers = F, testDispersion = F)
  474. ### Normality of random effects
  475. diag <- plot_model(model, type="diag")
  477. plot(diag[[2]]$Participant)
  478. plot(diag[[2]]$Word)
  479. plot(diag[[2]]$Channel)
  481. ### Homoskedasticity
  482. plot(diag[[4]])
  491. # LPP
  493. ## build and save model
  494. #model_full <- lmer(Amplitude_LPP ~ Coherence*Quotation*Pej.Value + (1 + Coherence*Quotation*Pej.Value|Participant) + (1 + Quotation*Pej.Value|Word) + (1 + Coherence*Quotation*Pej.Value|Channel),
  495. # data=LPP_df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  496. #save(model_full, file="LPP_model_full.rda")
  497. #load("LPP_model_full.rda")
  498. #summary(model_full)
  501. #model_nochslopes <- lmer(Amplitude_LPP ~ Coherence*Quotation*Pej.Value + (1 + Coherence*Quotation*Pej.Value||Participant) + (1 + Quotation*Pej.Value||Word) + (1|Channel),
  502. # data=LPP_df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  503. #save(model_nochslopes, file="LPP_model_nochslopes")
  504. #load("LPP_model_nochslopes.rda")
  505. #summary(model_nochslopes)
  507. #model_inter_nochslopes <- lmer(Amplitude_LPP ~ Coherence*Quotation*Pej.Value + (1 + Coherence:Quotation:Pej.Value|Participant) + (1 + Quotation:Pej.Value|Word) + (1|Channel),
  508. # data=LPP_df_long, verbose=1, control=lmerControl(optimizer="bobyqa", optCtrl=list(maxfun=100000)))
  509. #save(model_inter_nochslopes, file="LPP_model_inter_nochslopes.rda")
  510. #load("LPP_model_inter_nochslopes.rda")
  511. #summary(model_inter_nochslopes)
  513. #model <- model_inter_nochslopes
  515. #save(model, file="LPP_model.rda")
  517. ##load model
  519. load("LPP_model.rda")
  521. ## model summary
  522. model_sum <- summary(model)
  523. print(model_sum)
  524. write.csv(model_sum$coefficients, "./csvs/combined_model_summary_LPP.csv")
  527. ## estimated marginal means
  528. emmeans(model, pairwise ~ Quotation, adjust="holm")
  529. emmeans(model, pairwise ~ Coherence, adjust="holm")
  530. emtrends(model, pairwise ~ Coherence, var = "Pej.Value", adjust="holm")
  533. emm_means <- emmeans(model, ~ Coherence*Quotation)
  534. emm_means
  535. emm_contrasts <- contrast(emm_means, list(c(1, -1, 0, 0), c(0, 0, 1, -1), c(0, 1, 0, -1), c(1, 0, -1, 0)), adjust="holm")
  536. emm_contrasts
  539. emm_means_df <- as.data.frame(emm_means)
  540. emm_contrasts_df <- as.data.frame(emm_contrasts) %>% mutate(contrast = str_replace_all(contrast, "C ", "C-"))
  542. eta_squared(model, partial=TRUE)
  544. ## define plots variables
  545. means <- emm_means_df$emmean
  546. SEs <- emm_means_df$SE
  547. max_NoQM <- max((means+SEs)[1:2])
  548. max_QM <- max((means+SEs)[3:4])
  551. ## plot group differences
  552. QC_plot <- ggplot(emm_means_df, aes(x = Quotation, y = emmean, color = Coherence)) +
  553. ylim(-0.6, 0.6) +
  554. geom_point(position = position_dodge(width = 0.3), size=3) +
  555. geom_errorbar(aes(ymin = means-SEs, ymax = means+SEs),
  556. position = position_dodge(width = 0.3), # Shift error bars
  557. linewidth = 0.8, # Make error bars thicker
  558. width = 0.2) +
  559. labs(title = "",
  560. x = "Quotation",
  561. y = "LPP Amplitude (z-standardized)",
  562. color = "Coherence") +
  563. theme(text = element_text(size=14)) +
  564. geom_signif(y_position = c(max_QM+0.08, max_QM+0.08, max_QM+0.115, max_QM+0.25),
  565. xmin = c(0.925, 1.925, 1, 1),
  566. xmax = c(1.075, 2.075, 2, 2),
  567. annotation = c("", "", "*", "***"),
  568. tip_length = 0.1, color="black") +
  569. scale_color_brewer(palette="Dark2") +
  570. theme(legend.position = "bottom")
  573. plot(QC_plot)
  575. ggsave("./plots/QC_plot_LPP.png", dpi=300, width=1000, height=1200, units="px")
  578. ## plot three-way interactions
  579. all_plot <- plot_model(model, type="pred", terms=c("Pej.Value", "Coherence", "Quotation"), title="", axis.title=c("Pejorative Weight (z-standardized)", "LPP Amplitude (z-standardized)"), colors = "Dark2") +
  580. theme(legend.position = "bottom") + ylim(-0.6, 0.6)
  581. plot(all_plot)
  582. ggsave("./plots/all_plot_LPP.png", dpi=300, width=2000, height=1200, units="px")
  585. ## plot model estimates
  586. est_plot <- plot_model(model, type="est", colors="Set1", title="", axis.lim = c(-2.5, 2.5),
  587. axis.labels = c("Coh * Quot * Pej.Weight", "Quotation * Pejorative Weight",
  588. "Coherence * Pejorative Weight", "Coherence * Quotation",
  589. "Pejorative Weight", "Quotation", "Coherence"
  590. )) + theme(axis.text.y = element_text(angle = 30)) +
  591. ylim(-0.25, 0.25)
  593. plot(est_plot)
  594. ggsave("./plots/est_plot_LPP.png", dpi=300, width=1500, height=1800, units="px")
  597. ## combine plots
  598. combi_plot <- ggarrange(all_plot,
  599. ggarrange(QC_plot, est_plot, ncol=2, labels=c("B", "C")),
  600. nrow=2,
  601. labels="A")
  603. plot(combi_plot)
  605. ggsave("./plots/combi_plot_LPP.png", dpi=300, width=2400, height=2400, units="px")
  610. ## Diagnostics plots
  611. sim_res <- simulateResiduals(fittedModel = model, plot = F)
  613. ### Normality of residuals
  614. plotQQunif(sim_res, testUniformity = F, testOutliers = F, testDispersion = F)
  616. ### Normality of random effects
  617. diag <- plot_model(model, type="diag")
  619. plot(diag[[2]]$Participant)
  620. plot(diag[[2]]$Word)
  621. plot(diag[[2]]$Channel)
  623. ### Homoskedasticity
  624. plot(diag[[4]])

polpejEEG_repo.R, no license · at the source

Overview

Authors: Manuel Hons1, Edgar Onea2, Ilse Schiller1, Silvia Erika Kober1
  1. Department of Psychology, University of Graz, Graz, Austria
  2. Department of German Studies, University of Graz, Graz, Austria
Institutions: University of Graz (Austria)
Journal: Frontiers in human neuroscience, volume 20, article 1820376
Dates: received 1 March 2026; accepted 11 June 2026; published online 29 June 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.3389/fnhum.2026.1820376 · PMID 42444843 · PMCID PMC13357823 · OpenAlex W7166568562
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: EEG (modality), systems (subfield)
Methods: Preprocessing, Spectral & time-frequency, Statistics, Smoothing, state filtering, decompositions, Connectivity, Evoked potentials, Physiology & signal measures
Keywords: contextual violations, ERP, N400, political discourse, pragmatics
Topic: Neurobiology of Language and Bilingualism (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: Karl-Franzens-Universität Graz
Citations: not cited yet (Europe PMC); 66 references in the paper

Abstract

Introduction: Pragmatic violations are known to elicit reliable N400 modulations in event related potentials of electroencephalographic recordings. While this effect has been extensively documented in linguistic contexts, evidence from political discourse remains limited. To address this gap, the present study investigated neural responses to socio-pragmatic coherence in politically framed statements.

Methods: Participants read political utterances containing critical target words that were either coherent or incoherent with a preceding description of the producer. To examine whether pragmatic processing could be altered by typographic cues, target words were presented either unmarked or enclosed in quotation marks.

Results: Consistent with prior findings, pragmatically incoherent target words elicited a larger negative deflection peaking around 400 ms post-onset than coherent words. Notably, this pattern was reversed when target words were presented in quotation marks.

Discussion: We interpret this reversal as evidence that quotation marks signal a rejection or distancing from a word’s conventional meaning, thereby altering contextual integration demands. While rejecting the meaning of incoherent terms (e.g., “climate terrorists”) may facilitate integration when aligned with the speaker’s stance, rejecting the meaning of coherent terms (e.g., “climate activists”) appears to increase processing difficulty. These findings demonstrate that typographic cues can modulate socio-pragmatic interpretation in political language, with measurable neural consequences.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repository

Its files are read in the Code ↔ Paper reader above, with 2 matches between paragraphs and lines of code.

OSF sgbqf

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Languages: R (1)
Size: 7 files, 1 script
Software Heritage: not checked
Found in: the text, “Participants”
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: easystats (1 file), emmeans (1 file), ggplot2 (1 file), ggpubr (1 file), glmmTMB (1 file), lme4 (1 file), lmerTest (1 file), tidyverse (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
1 file
At the source: osf.io/sgbqf

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 1 script, each with its path and the digest of its content;
  • 2 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 2, 28 September 2026

  • Funding: added Karl-Franzens-Universität Graz

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, pages, dates, 4 authors, 5 keywords, 61 references.

Cite

This paper

Hons, M., Onea, E., Schiller, I., & Kober, S. E. (2026). The neural dynamics of political socio-pragmatic violations: an ERP study. Frontiers in human neuroscience, 20, 1820376. https://doi.org/10.3389/fnhum.2026.1820376

BibTeX

@article{hons2026neural,
author = {Hons, Manuel and Onea, Edgar and Schiller, Ilse and Kober, Silvia Erika},
title = {{The neural dynamics of political socio-pragmatic violations: an ERP study}},
journal = {Frontiers in human neuroscience},
year = {2026},
month = jun,
volume = {20},
pages = {1820376},
publisher = {Frontiers Media SA},
issn = {1662-5161},
doi = {10.3389/fnhum.2026.1820376},
url = {https://doi.org/10.3389/fnhum.2026.1820376},
pmid = {42444843},
pmcid = {PMC13357823}
}

RIS

TY - JOUR
AU - Hons, Manuel
AU - Onea, Edgar
AU - Schiller, Ilse
AU - Kober, Silvia Erika
TI - The neural dynamics of political socio-pragmatic violations: an ERP study
T2 - Frontiers in human neuroscience
J2 - Front Hum Neurosci
PY - 2026
DA - 2026/06/29
VL - 20
SP - 1820376
SN - 1662-5161
PB - Frontiers Media SA
DO - 10.3389/fnhum.2026.1820376
UR - https://doi.org/10.3389/fnhum.2026.1820376
LA - en
ER -

CSL-JSON

{
"id": "10.3389/fnhum.2026.1820376",
"type": "article-journal",
"title": "The neural dynamics of political socio-pragmatic violations: an ERP study",
"container-title": "Frontiers in human neuroscience",
"author": [
{
"family": "Hons",
"given": "Manuel"
},
{
"family": "Onea",
"given": "Edgar"
},
{
"family": "Schiller",
"given": "Ilse"
},
{
"family": "Kober",
"given": "Silvia Erika"
}
],
"container-title-short": "Front Hum Neurosci",
"volume": "20",
"page": "1820376",
"DOI": "10.3389/fnhum.2026.1820376",
"PMID": "42444843",
"PMCID": "PMC13357823",
"ISSN": "1662-5161",
"publisher": "Frontiers Media SA",
"URL": "https://doi.org/10.3389/fnhum.2026.1820376",
"language": "en",
"issued": {
"date-parts": [
[
2026,
6,
29
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1162/nol.a.264 [code]
The Temporal Dynamics of the Labeling Algorithm During Natural Language Comprehension: Neural Evidence for Phrase Grammatical Type Generation.
Journal: Neurobiology of language (Cambridge, Mass.)
In common: emmeans, lmerTest, lme4, EEG, 6 references
[2] doi:10.1371/journal.pone.0353990 [code]
Positive mood enhances accessibility of unrelated concepts in the first language but not in the foreign language.
Journal: PloS one
In common: emmeans, lmerTest, lme4, 3 other tools, EEG, 3 references
[3] doi:10.7554/elife.107088 [code]
Development of auditory and spontaneous movement responses to music over the first postnatal year.
Journal: eLife
In common: glmmTMB, easystats, emmeans, 4 other tools, EEG
[4] doi:10.1111/psyp.70306 [code]
From Feedback-Learning to Semantic Memory: Can Feedback-Related Brain Activity Predict Object-Word Associations?
Journal: Psychophysiology
In common: easystats, lmerTest, lme4, 3 other tools, EEG, 2 references
[5] doi:10.1016/j.neuroimage.2026.122115 [code]
Midfrontal theta power relates to response speeding following frustrative nonreward.
Journal: NeuroImage
In common: easystats, emmeans, lmerTest, 4 other tools, EEG, 1 reference
[6] doi:10.1111/psyp.70265 [code]
Neurocognitive Dynamics of Translating Information From a Spatial Map Into Action.
Journal: Psychophysiology
In common: easystats, emmeans, lmerTest, 4 other tools, EEG, 1 reference
[7] doi:10.1016/j.celrep.2026.117505 [code]
Impaired spatial coding and neuronal hyperactivity in the medial entorhinal cortex of aged APP knock-in mice.
Journal: Cell reports
In common: glmmTMB, easystats, emmeans, 4 other tools
[8] doi:10.1186/s13195-026-02036-1 [code]
Genetic drivers of progression in Alzheimer's disease are distinct from disease risk.
Journal: Alzheimer's research & therapy
In common: glmmTMB, easystats, emmeans, 4 other tools
[9] doi:10.1162/imag.a.1313 [code]
Functional specialization of angular gyrus and precuneus subregions for perspective-guided autobiographical memory retrieval.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: easystats, emmeans, lmerTest, 4 other tools, 1 reference
[10] doi:10.1093/braincomms/fcag217 [code]
Investigation of stress hormones across multiday seizure cycles.
Journal: Brain communications
In common: glmmTMB, easystats, lmerTest, 4 other tools

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.