R365 Day 1 “likert”

R365: Day 1 – mgcv likert
My home laptop is a Linux, and does not display all of the available packages (?). I am not sure if this means the other packages are not supported in Linux, or if I am currently unable to view them. The list of packages I can see consists of about 30 of the most powerful and useful packages (from what I can judge), rather than the 5000+ that are supported by R, presumably in Windows and/or Macintosh.
That being said, I scrolled down the list, closed my eyes, and picked the first package. mgcv. mgcv? Turns out it stands for “Mixed GAM Computation Vehicle with GCV/AIC/REML smoothness estimation”, which took me a minute to translate the acronyms, long enough to go “aaaaaahhhh” and rethink this whole blog. OR cheat and find a new package!
After a brief search, turns out you can install most packages (if you have root access in Linux) by just using install.packages(“[package name]”). This article from r-bloggers helped a lot (http://www.r-bloggers.com/installing-r-packages/). With this in hand, I searched through the comprehensive list of R packages and found the package “likert”, which helps to visualize likert type items. What in the world are likert items? After some quick Wikipedia-ing (a word which should be a verb, so I will treat it as such, suck it grammar nazis), I found out that I was already very familiar with the concepts of Likert Scales. If you have ever done a survey where the format is a statement (“I like rabbits”) followed by five options (strongly agree to strongly disagree), then you too have seen Likert Items Likert Scales. Rensis Likert (pronounced “LICK-urt”) was a psychologist from Michigan who worked on management theory. He developed his scale system as part of his PhD thesis in 1932 (always mind-boggling and a little depressing to find groundbreaking PhD theses). He found that the scales allowed for more exact results with fewer questions. A Likert Item is an individual statement and rating, whereas a Likert Scale is the sum of all of the ratings, and is used to compare across different groups.
So what does the likert package do? One of the major functions in the package likert is the function likert (go figure), which performs statistical tests on sets of Likert Items. Their example:
     library(“likert”)
 data(pisaitems)
    items29 <- pisaitems[,substr(names(pisaitems), 1,5) == ‘ST25Q’]
    names(items29) <- c(“Magazines”, “Comic books”, “Fiction”,
    “Non-fiction books”, “Newspapers”)
    l29 <- likert(items29)
    summary(l29)
    plot(l29)

plots responses to a survey about how often people read books and magazines. The summary statistics from likert describe the percentages of people in different groups, and the overall means (out of 5) for the different Likert Items. The package allows for Likert Items to be visualized as bar plots, heat maps, and histograms. The package is fairly new (October 2013), and seems to still  be partially in the works, but this would probably be a good tool for people who perform surveys and like data.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s