![]() ![]() To compare to a benchmark or to the populationĬomparing Lurie Children’s scores to that of a group of other children’s hospitals To compare two points in time, the same group of subjectsĬomparing the before and after scores of a group of children exposed to a certain treatment These are the tests we will cover in this post.Ĭomparing the scores of boys and girls who took the same test The table below gives you some basic guidelines for very common bivariate (two variable) tests. There is always a little confusion when you are beginning your statistical career about which test to use in which situation. This can become very onerous if you have multiple comparisons to run.Įxcel cannot run any multivariate statistics, so you will need to move to a different software package for regressions, MANOVAs and all of the other procedures that can be helpful to analysts. If you are comparing the means of two groups, you need to sort your data into those two groups and then highlight each groups range separately. You need to sort and copy the subset of data you want to a new sheet.Įxcel cannot 'break' your data for you. If you have some values you want to exclude from the analysis (like if you only want to compute the mean BMI on the teenagers in your sample), Excel does not give you a way to do that. So before you run any statistics, make sure your Excel sheet looks like a data file as much as possible. ![]() ![]() That can throw off the calculations considerably. What you need to be aware of is that if there is a missing row of data (a blank field), Excel usually reads that blank as a zero. Basic things Excel doesĮxcel uses ranges in its statistical functions, just like for all the rest of the computations it performs. I am going to assume you have read the post comparing the different kinds of variables and the table in the post that links those variables to the statistical procedures that are appropriate for them (see The thing of it all – on fields, variables and counting things) and that you know something about basic statistics (please see How Studying Statistics Guarantees a Happy Life and subsequent posts). So I thought it would be useful to show how to run a few basic statistical procedures and when to use them. You get a great deal of control and the program documents what you do, so you can always go back and replicate your work or find errors – really important things to achieving accuracy.īut the truth is, there are times that you just want to do something simple, and Excel can be a great tool. ![]() This is not true in software designed for statistics. It is also clunky – sometimes you have to do your own pre-calculations before running the statistical test. The challenge is that in Excel you do not get much control over the statistical tests you run and you cannot document well what you have done to prepare the data for analysis (data prep is nine-tenths of the law in statistics). It certainly represents the thinking of pretty much every person I have ever known who even dabbles in statistics.Įxcel is not evil or inaccurate. I happened to find this image online and laughed out loud when I read the caption. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |