User talk:Mw011235/sandbox

instructor review
This is looking really good! Very nice work. I like the example at the beginning. I do have a few suggestions:
 * 1. The last sentence in the first paragraph is really long - consider splitting it into two sentences
 * 2. Consider a one-sentence explanation of why one would run one sample tests (like the K-S test).
 * 3. More really long sentences under "Bivariate Analysis." Try to stick with one idea per sentence.
 * 4. Under "Linear Trends," perhaps it would be useful to describe how one obtains r?
 * 5. Data 'are' not 'is.' Check that the word 'data' is treated as plural throughout.
 * 6. Under "Visualization," do you plan to add examples? The narration is very detailed, but I find myself wanting to see an example.

peer review
Hi, guys, I just read your article, and these are my reviews: 1.The structure for the part 'Ways to analyze ordinal data' is not organized very well. For example, I cannot see clearly what are the ways used to analyze ordinal data. 2.In the sentence ‘Finding a mean or standard deviation for ordinal data is often discouraged, but other methods such as mean or mode should instead be used.’, I think there is a typo, the 2th ‘mean’ should be ‘median’.--XinmingLin (talk) 00:41, 21 February 2017 (UTC)

peer review
Hi guys, this is arman. I liked the way you structured your article's draft and the topic you want to work on. I would like to see some helpful examples so that readers can understand the whole concept of ordinal data. It must be a challenge to explain the ways ordinal data should be analyzed. Your resource citations seem reliable enough. In addition, you guys can think of adding several useful cool plots.

Armanaljahan (talk) 00:06, 23 February 2017 (UTC)

peer review
Hi guys, This is Yazhuo Deng. it is really great work! I like how you guys set the structure as well. The examples you provided are great. The article provide a lot of information on how ordinal data is similar to categorial data. It would be great if it can have some information about how can ordinal data be correctly treated as continuous variable. Because, often Likert scale variable is treated as continuous variable (of course, categorical variable can not be treated as continuous variable). — Preceding unsigned comment added by Deng9578 (talk • contribs) 05:35, 24 February 2017 (UTC)

Ordinal data article revision
Let's discuss what we should do to revise the article before moving it to the mainspace. Mw011235 (talk) 19:12, 24 February 2017 (UTC)

Nonparametric statistics
I'm thinking I should add some specific nonparametric tests to illustrate the range of what's available for ordinal data, along the lines of what you have done with the material from Agresti. That might help address Xinming's suggestion about clarification. Would that be an improvement? Mw011235 (talk) 19:12, 24 February 2017 (UTC)


 * I think that sounds great. I don't have anything on parametric statistics but I can find some if you want to throw a few things my way to write. Overall, I think having some of those tests will be useful. We can also add a "See also" page where there's a wiki page on kinds of categorical data analysis. Lond6846 (talk) 20:13, 24 February 2017 (UTC)


 * The "see also" idea is a good one. I'll add a subheading for nonparametric procedures below which I'll mention and link to some of the nonparametric procedures in Siegel & Castellan (1988). Mw011235 (talk) 21:41, 24 February 2017 (UTC)

Likert scale
I added the Likert scale example in table format, but I don't think it looks that great. Do you have any ideas on how to format it so it's "prettier"? Mw011235 (talk) 19:12, 24 February 2017 (UTC)


 * I don't know any formatting ideas, but I'll look around! Sorry!Lond6846 (talk) 20:13, 24 February 2017 (UTC)


 * I'll look around too. I've been meaning to get around to it but haven't had time yet. If I find anything, I'll let you know. Mw011235 (talk) 21:48, 24 February 2017 (UTC)

I like Yazhuo's idea of making an example of a Likert scale analyzed as a continuous variable. What do you think? I'm doing some brainstorming on that. Mw011235 (talk) 19:12, 24 February 2017 (UTC)
 * At some point, I did mention that ordinal data can be treated as a continuous variable in regression, but I'll have to look into it more. I am hesitant to automatically say "Yeah! Treat it as continuous!" since they do require different methods that continuous ones don't necessarily capture. Lond6846 (talk) 20:13, 24 February 2017 (UTC)


 * We could maybe connect this with the debate about using means, sd's, etc. for ordinal data in the articles I cited. I'll poke around a bit to see if I can come up with a good example. I just read a blog comment somewhere that indicated nonparametric and parametric procedures in the context of hypothesis testing with ordinal data generally come to the same conclusion. I have to wonder about that. And I'm not sure how one would think of a confidence interval in the context of an ordinal scale. Mw011235 (talk) 21:48, 24 February 2017 (UTC)

Data visualization
Do you have any ideas on plots or other data visualization for ordinal data, as suggested by Arman? I can only think of the standard approaches for categorical data (such as tables, pie charts, and bar charts), but maybe something has been developed for visualizing ordinal data that's not applicable to non-ordered categorical data? Mw011235 (talk) 19:12, 24 February 2017 (UTC)


 * I think that what you have listed there is about all there is for visualizing ordinal data, and I couldn't find anything else that would be special towards ordinal data. We can just let up links to other wiki pages, and maybe also talk about the proper way to design your bar chart. Lond6846 (talk) 20:13, 24 February 2017 (UTC)


 * Those sound like good ideas to me. Mw011235 (talk) 21:49, 24 February 2017 (UTC)


 * I found a few references for using color to represent ordinal data and added a bit of material from them to the data visualization section. It would be great if we could find or make some pictorial examples to include there. Mw011235 (talk) 20:10, 1 March 2017 (UTC)

some final edits
Phew, we're making it! I went through and changed a few of the big sentences and grammar issues. I also fixed up the linear trends (making that equation was bonkers!). For the visualization, Wikipedia can be pretty strict on what goes up. Any idea on how to get pictures that are allowed?Lond6846 (talk) 03:44, 7 March 2017 (UTC)


 * The edits you made definitely help the article's readability, and the additions to the linear trends section look really good. I'm playing around with a couple of example bar charts (one simple, one stacked) in R with toy data. I'll e-mail you the R code once they look OK (soon). Where can I find the Wikipedia restrictions on visual elements? Mw011235 (talk) 18:27, 7 March 2017 (UTC)


 * The page is Uploading images Lond6846 (talk) 17:55, 8 March 2017 (UTC)