sample_characterisation.Rmd

---
output:
  html_document:
    theme: spacelab
---

# Sample characterisation (age)

```{r, message = FALSE}
library("tidyverse")
library("rstatix")
library("ini")
```

Load the data from the big behavioural table:

```{r}
config <- read.ini("../config.ini")
tbl_path <- file.path(config$DEFAULT$QUESTIONNAIRE_DIR, "table.tsv")
df <- read_tsv(tbl_path, col_types = cols()) %>%
    select(c(subject_id:demonstrator_id, CONT_contingency, MFQ_ra, friendship_length))
```

Exclusions for technical reasons (3 subjects) and due to high number of motion outlier volumes (1 subject), all in the stranger group:

```{r}
df <- df %>%
	filter(! label %in% c("Trltdn", "Cgotop", "Plarre", "Nagery"))
```

As a sanity check, get the number of participants after those exclusions (note that the number of demonstrators has to be equal to the number of friend observers):

```{r}
df %>%
    count(role, group)
```

We are, however, interested only in the contingency-aware participants from now on (i.e. subjects for whom fMRI data is presented, not all who were recruited).

Age of the observers, regardless of group (note: age is the only numerical variable in `df` so no need to be explicit):

```{r}
df %>%
	filter(role == "OBS" & CONT_contingency == "YES") %>%
	get_summary_stats(age, show = c("mean", "sd", "median", "iqr", "min", "max"))
```

Age of the observers, broken down by group:

```{r}
df %>%
	filter(role == "OBS" & CONT_contingency == "YES") %>%
	group_by(group) %>%
	get_summary_stats(age, show = c("mean", "sd", "median", "iqr", "min", "max"))
```

For the demonstrators, drop those for whom neither friend nor stranger learned the contingency, and calculate the age statistics:

```{r}
dem_subset <- df %>%
    filter(role == "OBS" & CONT_contingency == "YES") %>%
    pull(demonstrator_id) %>%
    unique()

df %>%
    filter(role == "DEM" & subject_id %in% dem_subset) %>%
    get_summary_stats(age, show = c("mean", "sd", "median", "iqr", "min", "max"))
```

Length of the demonstrator-observer friendship (data collected only in the friend group. Both demonstrator and observer answered this question but we will stick to the observers' responses only):
* n=34 because subject Framwy did not respond

```{r}
df %>%
	filter(role == "OBS" & CONT_contingency == "YES" & group == "friend") %>%
	get_summary_stats(friendship_length,  show = c("mean", "sd", "median", "iqr", "min", "max"))
```

McGill Friendship Questionnaire score (Friend group only, observers):
* n=34 because subject Framwy did not respond

```{r}
df %>%
  filter(role == "OBS" & CONT_contingency == "YES" & group == "friend") %>%
	get_summary_stats(MFQ_ra,  show = c("mean", "sd", "median", "iqr", "min", "max"))
```

McGill Friendship Questionnaire score (Friend group only, demonstrators for whom observers learned contingency):
* n=34 because subject Framwy did not respond

```{r}
df_dem <- df
dem_subset <- df %>%
    filter(role == "OBS" & CONT_contingency == "YES", group == "friend") %>%
    pull(demonstrator_id) %>%
    unique()

df_dem %>%
    filter(role == "DEM" & subject_id %in% dem_subset) %>%
    get_summary_stats(MFQ_ra, show = c("mean", "sd", "median", "iqr", "min", "max"))
```