Skills Lab 05: Survey Data

class: center, middle, inverse, title-slide

.title[
# Skills Lab 05: Survey Data
]
.subtitle[
## <i>Ready, Steady, WRANGLE!</i>
]
.author[
### Dr Danielle Evans
]
.date[
### 25 February 2022
]

---

## Overview

- **Qualtrics**

+ What is it?
  + Setting up a survey
  + Navigating Qualtrics quirks
  + Downloading our data

- **R**

+ Reading in our data
  + Data exploration

- **Data Wrangling**

+ Recoding values
  + Fixing weirdness

--
<br>
<br>

.center[

###  *Let's get ready to wrangle!!*

]

---

## Qualtrics

- **Qualtrics** is an online survey software where you can create/host questionnaires and collect data from participants

- It's used widely by researchers in the school, and will be used by a lot of you for your **dissertations** next year (a scary thought I know...)

- It's relatively straightforward to use but there are some 'quirks' 🙄

- So first, we're going to run through setting up a survey in **Qualtrics** and how we navigate these 'quirks'

- Then I'll demo how we get these data into **R**, how we can explore our data, and how we might go about preparing our data for **factor analysis**

<br>

<div class="tu" style="font-size:90%">
<p><b>Top Tip!</b> Qualtrics can set stuff up weirdly - test it out thoroughly & check the data downloads appropriately <i>before</i> sending it out to participants! </p>
</div>

---

## Qualtrics

- **Qualtrics** is an online survey software where you can create/host questionnaires and collect data from participants

- It's used widely by researchers in the school, and will be used by a lot of you for your **dissertations** next year (a scary thought I know...)

- It's relatively straightforward to use but there are some 'quirks' 🙄

- So first, we're going to run through setting up a survey in **Qualtrics** and how we navigate these 'quirks'

- Then I'll demo how we get these data into **R**, how we can explore our data, and how we might go about preparing our data for **factor analysis**

<br>

<div class="pc" style="font-size:90%">
<p><b>Demo!</b> Setting up the survey! </p>
</div>

---

## Qualtrics Quirks

- We should check several parts of our **Qualtrics** survey *before* we send it out to participants:

- **Practical considerations**

+ Question wording 
  + Response type, requirements & validation
  + Coding/scoring of responses
  
- **Ethical considerations**

+ Participant information sheet
  + Informed consent
  + Debrief
  + IP addresses

<div class="swr" style="font-size:90%">
<p><b>Pet Peeve!</b> Qualtrics annoyingly collects <b>IP addresses</b> from our participants automatically - it's very important for <b>ethics</b> that we turn this off in settings!! </p>
</div>

---

## Qualtrics Quirks

- We should check several parts of our **Qualtrics** survey *before* we send it out to participants:

- **Practical considerations**

+ Question wording 
  + **Response type, requirements & validation**
  + **Coding/scoring of responses**
  
- **Ethical considerations**

+ Participant information sheet
  + Informed consent
  + Debrief
  + **IP addresses**
   
  
<div class="pc" style="font-size:90%">
<p><b>Demo!</b> Response types, response requirements & validation, survey workflow, changing item scoring, & removing IP addresses! </p>
</div>

---

## Reading & Exploring Data

- Once our survey is ready and has been filled in by participants, we can export our data file in a **.csv** format

- We've worked with these types of files LOADS so it isn't anything new to us

- We can use the **readr::read_csv()** function to import our data into **R**, in the same way we have done previously

- We then should employ a range of **functions()** to explore our data:

+ We want to check everything has imported as we expected it to
  + We want to figure out what data wrangling we need to do
  + We want to take note of any weirdness from our checks
  
<br>

<div class="pc" style="font-size:90%">
<p><b>Demo!</b> <b>Downloading our .csv, reading it into R</b> & exploring our data! </p>
</div>

---

## Data ExploRation

.center[

]

<div class="tu" style="font-size:90%">
<p><b>Top Tip!</b> Make notes and <b>#</b>comments as you explore & wrangle through your data, it really is super easy to forget what you've done & <i>why!</i> </p>
</div>

---
## Data ExploRation

.center[

]

<div class="pc" style="font-size:90%">
<p><b>Demo!</b> Downloading our .csv, reading it into R & <b>exploring our data!</b> </p>
</div>

---

## Keeping it simple

- When we download our data from **Qualtrics**, we'll often have a bunch of extra variables that we usually don't want or need

- These include things like: start date, end date, user language, duration, distribution channel...

- We might also get some extra rows that contain question labels or descriptions, and NOT data from our participants - we don't want this either!

- So, we can make our data easier to work with by using the **slice()**, **filter()**, and **select()** functions from **dplyr::**

- We can remove any extra rows, any variables we don't need, and any participants that didn't consent to our study

<br>

<div class="pc" style="font-size:90%">
<p><b>Demo!</b> Using <b>slice()</b>, <b>filter()</b> & <b>select()</b> to tidy up our data!</p>
</div>

---

## Help! I forgot to recode my items!

- After exploring your data, you might find that you didn't set up your survey properly

- But that's OK... we first need to figure out how the variables are currently measured/scored, and what they need to become

- Then we can wrangle our data and shape it into whatever we need using a variety of **functions()**

- The most useful functions for **recoding** come from **Base R** and the **dplyr**: 
  
  + **as.numeric()**
  + **replace()**
  + **dplyr::mutate()**
  + **dplyr::recode()**

<div class="mb" style="font-size:90%">
<p><b>Watch Out!</b> We don't <i>ever</i> want to make edits to our raw <b>.csv</b> file - we should only ever manipulate our data in <b>R</b>!! #openscience 🧠</p>
</div>

---

## Help! I forgot to recode my items!

**For recoding numeric data:**

.teeny[

```r
data <- data %>% dplyr::mutate(., col_name = dplyr::recode(col_name, `1` = 5))
```

**For recoding character data:**

```r
data <- data %>% dplyr::mutate(., col_name = dplyr::recode(col_name, "stduent" = "student"))
```

**For changing character data to numeric:**

```r
data <- data %>% dplyr::mutate(col_name = as.numeric(col_name))
```

**For recoding specific numeric values to NAs:**

```r
data <- data %>% dplyr::mutate(col_name = replace(col_name, col_name > 10, NA))
```

]

<br>

<div class="pc" style="font-size:90%">
<p><b>Demo!</b> Recoding numeric values, recoding text, recoding weiRdness to NAs</p>
</div>

---

## *That's all - happy wrangling!*

.center[

Give session feedback [here...](https://forms.gle/Yezaie6HukxT2cdM7) 😀

]