Task 1

data <- read.csv("sim_haywood_data.csv")

dim(data)
## [1] 400   5
str(data)
## 'data.frame':    400 obs. of  5 variables:
##  $ Participant_ID  : chr  "P31" "P79" "P51" "P14" ...
##  $ Prime_Type      : chr  "without_that’s" "without_that’s" "without_that’s" "without_that’s" ...
##  $ Context_Type    : chr  "Unambiguous" "Unambiguous" "Unambiguous" "Unambiguous" ...
##  $ Confederate_Type: chr  "Helpful" "Helpful" "Helpful" "Helpful" ...
##  $ Used_that_s     : int  1 0 0 0 0 0 1 0 0 0 ...
unique(data$Prime_Type)
## [1] "without_that’s" "with_that’s"
unique(data$Context_Type)
## [1] "Unambiguous" "Ambiguous"
unique(data$Confederate_Type)
## [1] "Helpful"
unique(data$Used_that_s)
## [1] 1 0
overall_proportion <- mean(data$Used_that_s)
overall_proportion
## [1] 0.3475

The dataset contained 400 observations (rows) and 5 variables (columns), with each row representing a single instruction given by one participant in one trial. Inspection of the variable types revealed that Participant_ID, Prime_Type, Context_Type, and Confederate_Type were categorical variables, while Used_that_s was a binary numeric variable (coded as 0 for absence and 1 for presence of “that’s”). The values matched the codebook specifications: Prime_Type consisted of two levels (“without_that’s” and “with_that’s”), Context_Type consisted of two levels (“Unambiguous” and “Ambiguous”), Confederate_Type consisted solely of “Helpful” (as expected for this simulated dataset), and Used_that_s contained only values of 0 and 1. The overall proportion of utterances in which participants used the disambiguating word “that’s” was 0.35 (34.75%), representing the marginal mean of the Used_that_s variable across all experimental conditions.

Task 2

stats <- data.frame(
  Prime_Type = c("without_that's", "without_that's", "with_that's", "with_that's"),
  Context_Type = c("Unambiguous", "Ambiguous", "Unambiguous", "Ambiguous"),
  N = c(
    sum(data$Prime_Type == "without_that’s" & data$Context_Type == "Unambiguous"),
    sum(data$Prime_Type == "without_that’s" & data$Context_Type == "Ambiguous"),
    sum(data$Prime_Type == "with_that’s" & data$Context_Type == "Unambiguous"),
    sum(data$Prime_Type == "with_that’s" & data$Context_Type == "Ambiguous")
  ),
  M = c(
    mean(data$Used_that_s[data$Prime_Type == "without_that’s" & data$Context_Type == "Unambiguous"]),
    mean(data$Used_that_s[data$Prime_Type == "without_that’s" & data$Context_Type == "Ambiguous"]),
    mean(data$Used_that_s[data$Prime_Type == "with_that’s" & data$Context_Type == "Unambiguous"]),
    mean(data$Used_that_s[data$Prime_Type == "with_that’s" & data$Context_Type == "Ambiguous"])
  ),
  SD = c(
    sd(data$Used_that_s[data$Prime_Type == "without_that’s" & data$Context_Type == "Unambiguous"]),
    sd(data$Used_that_s[data$Prime_Type == "without_that’s" & data$Context_Type == "Ambiguous"]),
    sd(data$Used_that_s[data$Prime_Type == "with_that’s" & data$Context_Type == "Unambiguous"]),
    sd(data$Used_that_s[data$Prime_Type == "with_that’s" & data$Context_Type == "Ambiguous"])
  )
)

stats$M <- round(stats$M, 2)
stats$SD <- round(stats$SD, 2)

knitr::kable(stats, caption = "Table 1. Means and Standard Deviations for Proportion of 'That's' Use by Prime and Context Condition")
Table 1. Means and Standard Deviations for Proportion of ‘That’s’ Use by Prime and Context Condition
Prime_Type Context_Type N M SD
without_that’s Unambiguous 100 0.12 0.33
without_that’s Ambiguous 100 0.25 0.44
with_that’s Unambiguous 100 0.53 0.50
with_that’s Ambiguous 100 0.49 0.50

Task 3

plot <- ggplot(stats, aes(x = Prime_Type, y = M, fill = Context_Type)) +
  geom_col(position = position_dodge(width = 0.9)) +
  geom_errorbar(aes(ymin = ifelse(M - SD < 0, 0, M - SD), 
                    ymax = ifelse(M + SD > 1, 1, M + SD)),
                position = position_dodge(0.9), 
                width = 0.25) +
  labs(
    x = "Syntactic Prime",
    y = "Proportion of Utterances with \"that's\"",
    fill = "Referential Context"
  ) +
  coord_cartesian(ylim = c(0, 1)) +
  theme_classic() +
  theme(
    legend.position = "right",
    axis.title = element_text(size = 12),
    axis.text = element_text(size = 10)
  )

print(plot)
<img src=“MR1_Kaja_files/figure-html/unnamed-chunk-3-1.png” alt=“Figure 1. Mean proportion of utterances containing”that’s” by prime condition and referential context. Error bars represent ±1 SD.” width=“672” />

Figure 1. Mean proportion of utterances containing “that’s” by prime condition and referential context. Error bars represent ±1 SD.

Task 4

Syntactic prime produced a much larger difference in “that’s” use than referential context. Participants who heard the confederate use “that’s” subsequently used it themselves approximately 51% of the time, whereas those who heard “without that’s” used it only 12% of the time. The effect of context was minimal and inconsistent across prime conditions. This pattern matches the findings of Haywood et al. (2005), who demonstrated that syntactic priming effects are substantially stronger than context effects in structural alternations.