```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```
+# Using built-in functions to conduct hypothesis tests
## T-tests
-You learned the theory/concepts behind t-tests last week, so here's a brief run-down on how to use built-in functions in R to conduct them and interpret the results.
+You learned the theory/concepts behind t-tests last week, so here's a brief run-down on how to use built-in functions in R to conduct them and interpret the results. As a reminder, t-tests are used to estimate the difference-in-means between two distributions. You might be working with a one-sample test, two (independent samples, pooled or paired samples, or difference scores. R can handle most of this stuff and, if not, remember that you know how to write your own functions!
+
+The usual procedure for conducting t-tests in R is pretty straightforward: you use the `t.test()` function. For my example, I'll simulate two groups by randomly sampling the `rivers` dataset. I will then conduct a two-sample t-test for difference of means under a null hypothesis of no difference.
+
+```{r}
+set.seed(20190503)
+group.a <- sample(rivers, 35, replace=TRUE)
+group.b <- sample(rivers, 42, replace=TRUE)
+
+t.test(group.a, group.b)
+```
+Notice everything that is contained in the output: It tells me what test I just ran (a two sample t-test). It also tells me the names of the two groups. Then it reports the test statistic (t) value, the degrees of freedom, and a p-value. It states my alternative hypothesis explicitly and includes a 95% confidence interval before also reporting the sample means for each group.
+
+Go read the documentation for `t.test()` in R as the function has many possible arguments to help you conduct exactly the sort of t-test you want given the particulars of your data, hypotheses, etc. The documentation will also tell you quite a lot about the output or "values" returned by the function. As with most tests and operations in R, you can store the output of a `t.test()` call as an object. In the code below, notice what values the object has and how you might use the object to call specific values later.
+
+```{r}
+t.result.ab <- t.test(group.a, group.b)
+
+t.result.ab ## same as before
+
+names(t.result.ab)
+
+t.result.ab$conf.int
+```
+You never need to calculate a confidence interval for a difference in means by hand again. You can even set other $\alpha$ values in the call to t-test if you want to estimate another interval (see the documentation).
## ANOVAs
-Analogous situation with t-tests. Here's a brief introduction to how they work in R.
+Here's a very brief introduction to how ANOVAs work in R. Let's use the `iris` dataset (look up `help(iris)` to learn about it) to build an example in which we test whether irises of different species have different average sepal width. Remember that the purpose of an ANOVA is to compare some continuous outcome across two or more categories to test the global hypothesis of any differences between groups.
+
+In the code block below, I explicitly store the `iris` dataset and then call `aov()` to conduct the test. In the call to `aov()` I indicate the variables and dataset I want to use. Notice the use of the tilde character ("~"). This is formula notation and tells R that I'm using the variables before and after as the dimensions of my ANOVA. The `aov()` function follows a convention that the dependent or outcome variable has to go first in the formula (just try it the other way if you want to see what happens — since ANOVAs assume a continuous outcome it won't work to try and use the categorical measure as the DV).
-## Visualizing confidence intervals
+```{r}
+iris <- iris
-We spent a lot of time on confidence intervals in the past few weeks. Since they can be so useful, surely we should learn some approaches to incorporating them into data visualizations.
+aov(Sepal.Width ~ Species, data=iris)
+```
+Wait a minute. That doesn't look quite right. Where's the F statistic? Where's the p-value? Why doesn't this table look like the ANOVA tables the book showed us?
+
+Not to worry. As with many statistical tests in R, the `aov()` function did everything it promised, you just need to ask it explicitly to show you the "summary" of the test to get the information you're looking for. You can also store the results and call `summary()` as a separate command.
+```{r}
+summary(aov(Sepal.Width ~ Species, data=iris))
+
+aov1.result <- aov(Sepal.Width ~ Species, data=iris)
+
+summary(aov1.result)
+```
+That should look more familiar. It's good to get accustomed to storing your model results and then calling `summary()` on the output. Such a routine is typical for many other types of statistical tests in R
+
+Note that the object returned by `aov()` has values that you can also call directly just like we did with the output of `t.test()` above. In the example below, I've asked R to show me the residuals (deviation from the sample mean across all groups) for each row of the dataset. I'm not sure what you might use this for, but it's available.
+
+```{r}
+names(aov1.result)
+
+aov1.result$residuals
+
+```
-## Date/time arithmetic
+As with all functions in R, I *strongly* encourage you to read the documentation before you set out to conduct your own ANOVAs in the wild. There are details and settings that it may be important to consider. In addition, you might notice in the documentation for ANOVAs that R actually conducts the ANOVA by performing a linear regression (whaaaat??!?!?). Turns out that ANOVA is just a special case of linear regression with some idiosyncratic assumptions and standards for the type of summary information that needs to be reported. No need to worry about the details of that for now, but it can be good to have in the back of your mind for future reference in case you find yourself in a situation where you had planned to run an ANOVA, but your data just doesn't fit the assumptions necessary to identify the model.
-Last, but not least, another wrinkle in time...or at least how to manage date-time objects in R.
\ No newline at end of file
projects / stats_class_2019.git / blobdiff