15  Lab 2 Checklist

What you should have when Clean Data is done

Abstract

This short chapter closes Lab 2. It first sums up each Clean Data chapter in three parts: what you learned in The Play, what you did in The Lab (and whether you did it in Qualtrics with your team or in RStudio on your own computer), and what you will do in Your Turn when you come back with your team’s data. It then lists what should exist in your team’s Google Drive, on your computer, and in your lab file, with a one-line way to check each item and a link back to the section that made it. Lab 2 has two halves: the team work in Qualtrics (naming, recoding, codebook, export), which has no R in it, and your first analysis in RStudio, importing and cleaning the lab dataset and building the categorical and composite variables that every later lab uses.

Keywords

checklist, Lab 2, clean data

Lab 2 is the first lab with code in RStudio, and it is also the only lab that is half team work. The first two chapters (Name and Recode Variables in Qualtrics and Export Survey Data) happen in Qualtrics with your team; the rest happen in your own RStudio with the lab dataset. The list below is split the same way. (If you are not sure what a Lab checklist is for, Lab 1 Checklist explains the two kinds.)

15.1 What each chapter asked of you

Every row below has the same three parts as the chapters themselves. The Play is what you read: the chapter walks through the code with an example dataset, and you do not retype it. The Lab is what you did yourself. Your Turn is what you will do later with your team’s data, once data collection has closed.

The Where column is about The Lab, because that is where students lose track of whether they were supposed to be coding:

• Qualtrics, with your team: no R at all. Your team’s own survey is the practice.
• RStudio, lab data: you typed and ran code in Lastname_lab2.qmd on your own computer, using the lab dataset from the class Google Drive, and checked your answers against Ref’s checks.

Chapter	The Play: what you learned	The Lab: what you did	Where	Your Turn: with your data
Name & Recode Vars	The naming rules (NAME1, _01, _QUAL), a number for every answer choice, and what a codebook is for	Named and recoded every question in your team’s survey, filled in the Team Codebook, and exported the draft survey questions to Word	Qualtrics, with your team	Export the final survey questions and codebook before the survey is published; after that, only the changes listed as safe
Export Survey Data	The export settings, the .cleandata copy, and the file name	Triple-checked the coded values and exported the final survey questions. You did not export data yet: your team is still collecting it	Qualtrics, with your team	After data collection closes: export the .xlsx, make the .cleandata copy, name it MONTH.DAY.YEAR.team#.cleandata.xlsx, and meet with a Quant Peer Mentor
Import Data Once	How to read an import line, how to look at what came in (head(), names(), nrow(), ncol()), and that only the function changes between .csv and .xlsx	Lab Plays 1 to 5: imported the lab data as mh_raw, compared the names to the codebook, turned -99 and -50 into NA, and fixed a number column that came in as text. This is the first code you ran in RStudio on data	RStudio, lab data	Import your team’s .cleandata file as alldata, with the same checks
Pre/Post Data (Teams 3 and 5 only)	Long and wide data, and how PASSWORD links a person’s two responses	Lab Plays 1 to 3: imported the long lab file, cleaned the password, found the duplicate, and reshaped to wide	RStudio, lab data	Template 1 on alldata: one row per person, with _PRE and _POST columns
Vignette Data (Team 4 only)	How a two-by-two design becomes a table, one row per rating, and a factor for each thing that varied	Lab Plays 1 to 3: reshaped CASE1 to CASE4 to long, added the two factors, and found the four cell means	RStudio, lab data	The same template on SENTENCE1 to SENTENCE4
Mutate Variables	Three moves: collapse categories, cut a score into bands, combine select-all boxes; and the _#CAT naming rule	Lab Plays 1 to 4: ran source("lab_prep.R") to get mh_clean, then built EDUCATION_3CAT, POL_3CAT, HEALTH_3CAT, and SES_01, counting each one	RStudio, lab data	Make every categorical variable your report uses from alldata, count each one, and cite any cut-off
Creating Composites	Scoring keys, reverse items, and Cronbach’s alpha	Lab Plays 1 to 4: scored WELLBEING, STIGMA with its two subscales, and the K10 with scoreItems(), then wrote a Measures sentence for each	RStudio, lab data	Score every multi-item measure in your team’s survey and report its alpha

ImportantImportant: The Lab is practice on the lab dataset. Your Turn is not due yet

In Lab 2 you do The Lab of each chapter that applies to you, in Lastname_lab2.qmd. You do not do the Your Turn sections now, because your team’s data do not exist yet. When they do, you will open each of these chapters a second time, skip to Your Turn, and run the template on alldata in your report file. The last column of the table is your list for that day.

15.2 Part 1: your team, in Qualtrics and the Google Drive

These are team items. One person can check them for everyone, but every team member should know where the files are.

☐	What	How to check	Where it came from
☐	Every variable has an export tag following the rules	Open the survey; no question is still named Q1, Q2…; multi-item measures are NAME1, NAME2…; yes/no variables end in _01; open-ended questions end in _QUAL	Name and Recode Variables, Name Your Variables in Qualtrics
☐	The shared class variables are present and named exactly	CONSENT, GENDER (0 to 3), RACIALIZED (select-all), and TIME for Teams 3 and 5	Name and Recode Variables, the shared-variables table
☐	Every answer choice has a number	Recode values are on; scales start at 1, true zeros at 0, every binary is No = 0 / Yes = 1, -99 = Prefer not to say, -50 = Don’t know	Name and Recode Variables, Recode the Values
☐	The Team Codebook is filled in	TeamCodebook.docx in the team Google Drive lists every variable with its values	Name and Recode Variables, Create a Team Codebook
☐	The survey questions are exported to Word with coded values showing	Draft.TeamSurveyQuestions.docx (and later the final version) in the team Google Drive	Name and Recode Variables, Export Survey Questions
☐	You know how the data export will be done	Excel, numeric values, “Recode seen but unanswered as -99” checked, split multi-value fields for select-all questions, the .cleandata copy with the extra header rows removed, and the file name MONTH.DAY.YEAR.team#.cleandata.xlsx	Export Survey Data, Export Survey Data (.xlsx) and Rename Files

GoGo: your team’s data file is not part of Lab 2

Your team is still collecting data during Lab 2. The R half of this lab uses the lab dataset, which is why every chapter has a Lab (lab data) and a Your Turn (team data). You will come back to the Your Turn sections with your team’s .cleandata file after data collection closes, with a Quant Peer Mentor (Caution: meet with a Quant Peer Mentor).

15.3 Part 2: you, in RStudio

☐	What	How to check	Where it came from
☐	Lastname_lab2.qmd exists in your project with a YAML header and a load-library chunk	The Files tab shows it next to Lastname_FRI.Rproj; the load-library chunk loads tidyverse, readxl, and psych without error	Start a Report, Name the file
☐	The lab dataset is in data/	list.files("data") shows ANTH306_LayConceptionsMH_SYNTHETIC.xlsx (and, for Teams 3 and 5, ANTH306_LayConceptionsMH_SYNTHETIC_LONG.xlsx)	Import Data Once, Before you start, and the box Where the data files are
☐	lab_prep.R is in the project folder, not in data/	list.files() shows lab_prep.R next to Lastname_FRI.Rproj; source("lab_prep.R") runs silently and nrow(mh_clean) prints 188	Transforming Your Data, Important: where mh_clean comes from
☐	You imported the lab data yourself as mh_raw and looked at it	nrow(mh_raw) is 200 and ncol(mh_raw) is 134; you ran head() and names()	Import Data Once, Lab Plays 1 and 2
☐	You checked the names against the codebook with setdiff() in both directions	You can say which variables are in the data but not the codebook, and why	Import Data Once, Lab Play 3
☐	You turned -99 and -50 into NA	summary(mh_raw$POLITICALBELIEFS) shows a minimum of 1 and an NA's count, not -99	Import Data Once, Lab Play 4
☐	You converted a text column to numbers with as.numeric() and counted what became NA	veggiescore_SYNTHETIC.xlsx is in data/; summary() of VEGGIESCORE shows a minimum and maximum and 24 NA's	Import Data Once, Lab Play 5
☐	Teams 3 and 5 only: you reshaped the long lab file to wide	lab_wide has one row per PASSWORD with HELPSEEK_PRE and HELPSEEK_POST	Pre/Post Data, Lab Plays 1 to 3
☐	Team 4 only: you reshaped the vignette ratings to long and added the two factors	One row per person per case (CASE, RISK), with the CASE_RACIALIZED and CASE_SEVERITY factors, and four cell means	Vignette Data, Lab Plays 1 to 3
☐	You built the four lab categories with mutate() and case_when()	count() on EDUCATION_3CAT, POL_3CAT, HEALTH_3CAT, and SES_01 matches the Ref’s checks	Transforming Your Data, Lab Plays 1 to 4
☐	You scored three composites with scoreItems(impute = "none") and read their alphas	WELLBEING, STIGMA (with its two subscales), and the K10 total; the alphas match the Ref’s checks	Creating Composite Scores, Lab Plays 1 to 3
☐	You wrote one Measures sentence per composite	Scale, number of items, response scale, scoring, reverse items, alpha in this sample	Creating Composite Scores, Lab Play 4
☐	Lastname_lab2.qmd renders and is backed up	Render produces Lastname_lab2.html with no red text; a copy of the .qmd is in the R Labs folder of your ELN	Find the Ref, Ref’s final whistle

Caution Ref’s call: the gaps that break Lab 3 and Lab 4

• object 'mh_clean' not found in a later chapter → source("lab_prep.R") was not run in this session, or the file is inside data/ instead of the project folder.
• object 'WELLBEING' not found → same cause; every composite and every _CAT variable the later labs use is created by lab_prep.R, not by you typing it again.
• Your lab file lives on the Desktop → relative paths fail. It belongs in Lastname_FRI.
• You skipped Lab Play 3 (the setdiff() check) → nothing breaks in the lab data, which is clean, but the same check is the one that catches a misnamed variable in your team’s export, where it matters.

15.4 What Lab 3 will add

Lab 3 has very little code. You will read the Analysis Map, take a short quiz placing five research questions on it, check whether an outcome is normal, and write the Data Analysis Plan paragraph for your Methods section. Lab 3 Checklist is the matching list.