Dev 2021

[lbednarczyk]

No description provided.

[JingFu1]

Recommended Priority: Urgent

According to the style guide, I think "ctb" is better than "aut"

[sv101]

Looks good!

[neilhatfield]

I have a few changes I would like you to make and some things to investigate. On the whole the app appears to be in pretty decent shape.

DESCRIPTION

• Change the title to Analysis of Covariance
• Change the date to 2018-08-01
• Change the Lifecycle to questioning
• Change the Learning Outcome to "The student will learn to use interaction plots to decide whether interaction terms are significant in ANCOVA settings."

README.md

• Spell out ANCOVA in the title
• Add "# App Description" above the description that is there.

Files

• Delete model2.csv as this does not appear to be used.
• Delete all image files from the www folder which are no longer used. For example, arrow.gif, arrow.png, check.png, etc.
• Delete the navcolor.css file from the www folder

app.R

• Delete library(fontawesome) call and update the reference section accordingly.
• Uncomment out library(sortable) as this package is needed
• Add a section header for the game page in the UI definition (i.e., "### Game page ----")
• Move the code that loads the data sets from between the UI and server to the labelled section just above the UI
• Fix the formatting of the sendSweetAlert for the info button (remove the colon and add type = "info")
• Change all instances of snake_case to camelCase to be aligned with the Style Guide
• Check spelling and grammar. (E.g., "continous" should be "continuous"; "Select covariance" should be "Select covariate")
• Change any tags$b to tags$strong
• Put one space on both sides of symbols such as =, ==, and <-

Things to Investigate and Report Back On

• I don't understand the presence of the Diagnostic Plots subsection of the Prereq's. I didn't see where those diagnostic plots are getting used elsewhere in the app. Did I miss something? If such plots aren't being used, then we should remove them from the prereq's
• Currently, the app displays raw R output for the anova table. Should we convert this to a more polished version (like what I require in class)?
• Why does the app fit models for the otter and diet data sets and then use predicted data for making the plots? Why aren't we just using the actual data for the plots?
• Where does the output$p go? I can't find a corresponding UI part. Is this something that we need to add to the UI somewhere?
• What is the role of maxScore and the commented out code towards the end of the server?

[luqije]

Report my reviews for "Things to Investigate and Report Bacnk On":

• The diagnostic plots are not necessarily needed in Pre. Because later in the explore page and game page, we are not showing information related to these. Also, the app "Assumptions of ANOVA" offers information related to ANCOVA.
• We should convert the R output to a polished version. Since not every user uses R when analyzing ANCOVA. Displaying the output using the ds version (used in the paper) will be better.
• I think using a random dataset on Explore page can help the users more directly to see how things changes can affect the interaction with the p-value through the plot. However, in this case, we can add information about the meaning behind them. By practicing this, the users can explore the matching game on the Game page.
• output$p is missing in the UI part of the Explore page. I believe the author is trying to show the feedback of what is the meaning of the p-value of each output with different datasets.
• maxScore is missing in the UI part of the Game page. I think it's meant for calculating the total score of each trial.

Error and Suggestions:

• The plots for the Diet dataset when choosing the variable Age with Gender and Pre-diet Weight with Gender are not showing with errors.
• Add brake between each explanation in Pre would be helpful.

[Sburke26]

Here is what I noticed while reviewing:

There appears to be two README files.
The year badge in the README isn't updated properly.
The info pop-up does not have a circled i icon.
The app does not run with the object named "abChange". I had to change it to "ab_change" for it to run.
The prerequisites page could be improved to be more attractive and beneficial to the user.
In particular, removing the diagnostic plots is one way, as it appears they don't support what is covered in the app. Also, maybe switching to a collapsible tab for each type of analysis instead of a spaced out table would be better. When the window of the app is minimized, this spaced out table is hard to look at.
There also appears to be a few instances of commented out code. Some of which may come in handy if the app were to try to implement certain features (ex.score,).
There are also some issues with spacing in the code.

Edit:
Also I noticed there was a console message: "Warning in is.na(x) : is.na() applied to non-(list or vector) of type 'closure'".