In my view, a good graph provides a reasonably accurate picture of the main patterns in the data and of how the raw data relate to these patterns – i.e. is there a lot of variation or do the individual data points map closely onto the patterns? This makes graphs – rather than numerical descriptions or significance tests – essential for presenting research results to an audience, especially one that may not be familiar with advanced statistical techniques or even one that may not be entirely comfortable with concepts such as, say, standard deviations or confidence intervals (any casual definition of either of which is almost certainly wrong). Second, while a good graph can be difficult to construct, it should – by virtue of being a good graph – be straightforward to comprehend with little guidance on the part of the author or presenter. This is unfortunate, as a good graph serves two important purposes:įirst, it can alert the researcher to aspects of the data that aren’t obvious from a purely numerical description, such as outliers, coding errors, non-linearities, and skewed distributions. When flicking through an issue of a journal on language research or when attending a conference, chances are you’ll harvest a fair number of unclear, uninformative, for-the-record-only graphs. It is geared towards readers who don’t have much experience with drawing statistical graphics and who aren’t entirely happy with their attempts in Excel. This blog post is a step-by-step guide to drawing scatterplots with non-linear trend lines in R.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |