> 02 🤝
> say hello to your data

🔗 robust-tools.djnavarro.net

danielle navarro

👩‍🔬

Hayes, Banner, Forrester & Navarro (2019).

Selective sampling and inductive inference: Drawing inferences based on observed and missing evidence

https://psyarxiv.com/2m83v/

https://robust-tools.djnavarro.net/reasoning/

Property sampling: the robot only detects plaxium spheres

Category sampling: the robot only tests small spheres

Small sample size: Elicit judgments after two observations

Medium sample size: Elicit judgments after six observations

Large sample size: Elicit judgments after twelve observations

Seven test items that vary in size: Smallest...

Seven test items that vary in size: Largest...

smooth generalisation profiles
similar for small samples
different for large samples
property shows "tightening"
category does not

Exercise #1:
discuss this study!

01:00

📖
Reading
your data

https://rstudio.cloud/project/978818

Reading the data into R

library(tidyverse)
frames <- read_csv(file = "data_reasoning.csv")

Reading the data into R

library(tidyverse)
frames <- read_csv(file = "data_reasoning.csv")

## Parsed with column specification:
## cols(
##   id = col_double(),
##   gender = col_character(),
##   age = col_double(),
##   condition = col_character(),
##   sample_size = col_character(),
##   n_obs = col_double(),
##   test_item = col_double(),
##   response = col_double()
## )

Click on "frames" in the environment pane

Inspecting the data

print(frames)

Inspecting the data

print(frames)

## # A tibble: 4,725 x 8
##      id gender   age condition sample_size n_obs test_item
##   <dbl> <chr>  <dbl> <chr>     <chr>       <dbl>     <dbl>
## 1     1 male      36 category  small           2         1
## 2     1 male      36 category  small           2         2
## 3     1 male      36 category  small           2         3
## 4     1 male      36 category  small           2         4
## 5     1 male      36 category  small           2         5
## 6     1 male      36 category  small           2         6
## # … with 4,719 more rows, and 1 more variable:
## #   response <dbl>

Inspecting the data

glimpse(frames)

Inspecting the data

glimpse(frames)

## Rows: 4,725
## Columns: 8
## $ id          <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ gender      <chr> "male", "male", "male", "male", "mal…
## $ age         <dbl> 36, 36, 36, 36, 36, 36, 36, 36, 36, …
## $ condition   <chr> "category", "category", "category", …
## $ sample_size <chr> "small", "small", "small", "small", …
## $ n_obs       <dbl> 2, 2, 2, 2, 2, 2, 2, 6, 6, 6, 6, 6, …
## $ test_item   <dbl> 1, 2, 3, 4, 5, 6, 7, 1, 2, 3, 4, 5, …
## $ response    <dbl> 8, 7, 6, 6, 5, 6, 3, 9, 7, 5, 6, 4, …

Exercise #2

The pipe
%>%

The pipe, %>%

Take the frames data...
Do one thing...
Then another...
And then one more...

The pipe, %>%

Take the frames data...
Do one thing...
Then another...
And then one more...

frames %>%
  do_one_thing(.) %>%
  then_another(.) %>%
  and_then_one_more(.)

robust-tools.djnavarro.net

data %>% 
  tidy() %>%
  describe() %>%
  visualise() %>%
  analyse()

We "pipe" our data through operations
The data set flows through our analysis

   
`

Exercise #3

Grouped
summary

The basic idea

frames %>%
  group_by( GROUP ) %>%
  summarise( EXPRESSION ) %>%
  ungroup()

use group_by to define groups
use summarise to... summarise
use ungroup to remove grouping

Group, summarise...

frames %>%
  group_by(test_item, sample_size, n_obs, condition) %>%
  summarise(response = mean(response))

Group, summarise...

frames %>%
  group_by(test_item, sample_size, n_obs, condition) %>%
  summarise(response = mean(response))

## # A tibble: 42 x 5
## # Groups:   test_item, sample_size, n_obs [21]
##   test_item sample_size n_obs condition response
##       <dbl> <chr>       <dbl> <chr>        <dbl>
## 1         1 large          12 category      7.60
## 2         1 large          12 property      7.16
## 3         1 medium          6 category      7.32
## 4         1 medium          6 property      6.66
## # … with 38 more rows

Group, summarise, and ungroup

frames %>%
  group_by(test_item, sample_size, n_obs, condition) %>%
  summarise(response = mean(response)) %>%
  ungroup()

## # A tibble: 42 x 5
##   test_item sample_size n_obs condition response
##       <dbl> <chr>       <dbl> <chr>        <dbl>
## 1         1 large          12 category      7.60
## 2         1 large          12 property      7.16
## 3         1 medium          6 category      7.32
## 4         1 medium          6 property      6.66
## # … with 38 more rows

A more realistic example

frames %>% 
  group_by(test_item) %>%
  summarise(
    m = mean(response),
    s = sd(response),
    n = n()
  ) %>%
  ungroup()

A more realistic example

frames %>% 
  group_by(test_item) %>%
  summarise(
    m = mean(response),
    s = sd(response),
    n = n()
  ) %>%
  ungroup()

## # A tibble: 7 x 4
##   test_item     m     s     n
## *     <dbl> <dbl> <dbl> <int>
## 1         1  6.77  2.56   675
## 2         2  6.88  2.10   675
## 3         3  5.71  2.41   675
## 4         4  4.48  2.68   675
## 5         5  3.76  2.81   675
## 6         6  3.43  2.99   675
## 7         7  3.26  3.11   675

these are bad variable names
i've done it to fit the slides
don't do it in real life

A more realistic example

by_item <- frames %>%
  group_by(test_item) %>%
  summarise(
    m = mean(response), 
    s = sd(response), 
    n = n() 
  ) %>%
  ungroup()

A more realistic example

ggplot(
  data = by_item,
  mapping = aes(
    x = test_item, 
    y = m
  )
) +
geom_point()

A more realistic example

ggplot(
  data = by_item,
  mapping = aes(
    x = test_item, 
    y = m
  )
) +
geom_point()

A more realistic example

ggplot(
  data = by_item,
  mapping = aes(
    x = test_item, 
    y = m, 
    ymin = m - s,
    ymax = m + s
  )
) +
geom_pointrange()

Exercises #4 and #5

🖋
Writing
data

If we read with read_csv...

read_csv("data_reasoning.csv")

If we read with read_csv...

read_csv("data_reasoning.csv")

... then we write with write_csv!

write_csv(by_item, "summary_reasoning_by_item.csv")

Exercises #6 and #7

Where
next?

Not in this class, but for future reference...

Excel files? readxl package
SPSS, Stata or SAS? haven package
JSON format? jsonlite package
Databases? dbplyr package

Want more information?

Data import chapter in R for Data Science
https://r4ds.had.co.nz/data-import.html

↑, ←, Pg Up, k	Go to previous slide
↓, →, Pg Dn, Space, j	Go to next slide
Home	Go to first slide
End	Go to last slide
Number + Return	Go to specific slide
b / m / f	Toggle blackout / mirrored / fullscreen mode
c	Clone slideshow
p	Toggle presenter mode
t	Restart the presentation timer
?, h	Toggle this help

> 02 🤝 > say hello to your data

🔗 robust-tools.djnavarro.net

danielle navarro

https://robust-tools.djnavarro.net/reasoning/

https://rstudio.cloud/project/978818

Help

> 02 🤝
> say hello to your data