This preview shows page 1. Sign up to view the full content.
Unformatted text preview: y/quantile plot
up with an error message. Try the following:
Boxplot with outliers*
1. Press E and clear any extraneous equations or
keystrokes by scrolling down and pressing P.
*The boxplot not described above treats all data points as nonoutliers.
2. If you have more than one stat plot on, you may need to
3. To select the plot type, use the arrow keys to move around.
try plotting the plots one at a time. Not all plot types are
Here, the boxplot with outliers option has been selected.
compatible.
Note, you must press [ to select a type.
3. If you are making a scatterplot, your lists must be the same
4. Xlist signiﬁes the list containing the data you want to look at.
length—check that your explanatory and response variables
To select a list, highlight the list name (in this case, L1),
have the same number of values. To ﬁnd the DISTR (distribution) menu, press F O. The screenshots to the right all show calculations
dealing with the normal curve with mean 100 and standard deviation 15.
To find the area under the normal curve between two given values of your variable:
1. From the DSTR menu, select option 2:normalcdf(.
2. The normalcdf() command requires four inputs, separated by commas. You will need to ﬁnd and use the
, key. Those four inputs are, in order, the left bound of the region, the right bound of the region, the
mean of the distribution, and the standard deviation. Press [.
To find the area under the normal curve to the left or right of a particular value of your variable:
1. From the DSTR menu, select option 2:normalcdf(.
2. To ﬁnd the area to the left: Because the left bound is (theoretically) q , we make up a large negative
number, such as –9999, to stand in for it. The remaining inputs to normalcdf() are as before: separated
by commas, we type in the right bound of the region, the mean, and standard deviation. Press [.
3. To ﬁnd the area to the right: First, type in the left bound of the region. As before, we make up a large
positive number to stand in for the right bound, such as 9999, and then continue with the mean and
standard deviation, all separated by commas. Press [.
To find the variable value associated with a given percentile:
1. From the DSTR menu, select option 3:invNorm(.
2. The command invNorm() requires three inputs, separated by commas. These three inputs are, in order,
the percentile in decimal form, the mean, and the standard deviation of the distribution. Press [.
The percentile can also be interpreted as the area to the left of the answer found by invNorm(). Scatterplots, Correlation, Finding the Regression Statistics, and Residual Plots (continued)
To calculate correlation and the other regression statistics:
1. From the STAT menus, choose CALC. Scroll down to choose either option 4:LinReg(ax + b) or option
8:LinReg(ax + b). Your text or instructor may recommend one over the other. Note the different roles
the parameters a and b play in each command.
2. The LinReg commands expect at least two inputs, the explanatory list name and the respo...
View
Full
Document
This document was uploaded on 01/28/2014.
 Spring '14

Click to edit the document details