This preview shows page 1. Sign up to view the full content.
Unformatted text preview: nse variable
list name, in that order, separated by a comma. Press [ to see the results.
3. You should see values for four statistics on your screen. If not, see the Note below. For option 8, the
statistics a and b are the yintercept and slope of the regression line. For option 4, the statistics a and b
are the slope and yintercept. The statistics r2 and r are the Rsquared statistic (also called the coefﬁcient
of determination) and correlation, respectively.
Note: If you are using this feature for the ﬁrst time or have recently changed the batteries in your calculator
you may see the results only for the a and b. To see the correlation and Rsquared, you must do the following:
1. Press F and 0 to see the CATALOG of commands stored in your calculator.
2. Either scroll down to the command DiagnosticOn or press the alpha key D to get there faster.
Select DiagnosticOn and press [ until you see the Done on your screen.
3. Redo the LinReg command to ﬁnd the correlation.
To make a plot of the residuals versus the explanatory variable:
1. After performing a LinReg command, turn on one plot from the STAT PLOT menu.
2. For Type, choose a scatterplot, which is the very ﬁrst option, and press [.
3. Xlist will still contain the list name of your explanatory variable, L1. To type in L1, choose F and
the number 1.
4. Ylist will contain the residuals. With the cursor in the Ylist line, choose F K to select the LIST
menu. Scroll down and highlight RESID. Press [. Your screen should look like the one to the
right. To display the plot, remember to press B 9. Generating Random Integers
1.
2. Assessing Normality Press L and scroll to the PRB menu. Choose option 5:randInt(.
The command randInt() can take up to three inputs. The ﬁrst two are the boundary numbers. The last
input (optional) is the number of these numbers you want to generate. Press [. Here, four integers
between 1 and 6 are generated. By pressing [ twice, two more random lists of these integers are
generated.
You can store a large number of these random integers in a list. Before pressing [ in step 2, press
Y and select a list name. Press [. The easiest way to view these numbers is to go to the lists
via K and select 1:Edit. Often, in statistical analyses, we are required to verify that our data plausibly come from a normal distribution.
We can review a histogram of our sample to see that the data have a roughly bellshaped form. A more thorough
way to check normality is to make a special plot called a normal probability plot (sometimes called a normal
quantile plot).
1. From the STAT PLOTS menu, turn on one plot.
2. For Type, choose the very last option, as in the screen to the right. Press [.
3. To display the plot, remember to press B 9. 3. Descriptive Statistics: Summarizing Quantitative Variables Scatterplots, Correlation, Finding the Regression Statistics, and Residual Plots Press K and select CALC. Select option 1:1–Var Stats.
Type in your list name by pressing F and one of the number keys from 1 through 6. Press...
View
Full
Document
This document was uploaded on 01/28/2014.
 Spring '14

Click to edit the document details