Sie sind auf Seite 1von 5

Outliers and Influential Points

Statistical Analysis 9
Winter 2016-17

Weve talked about

1. Open TI-Nspire. Use the boats vs manatee data I have collected for you.

a. In TI-Nspire, make a boxplot for each variable. (Insert > Data & Statistics >
Add X Variable > Right click on whitespace > Boxplot). Take a screenshot, and insert it
below.

b. Make a scatterplot. (Insert > Data & Statistics > Add X Variable > Add Y
Variable). Take a screenshot, and insert it below.
c. Here is the spot: What do you notice? What do you wonder?

There is a strong correlation between the number of registered boats and the number of manatees
killed.

There are 0 outliers.

d. Now, in TI-Nspire, go to the window that has your scatterplot. Try to fit a
line to your data! To do this, go to Tools > Analyze > Regression. Choose the one that
fits your data the best (you may need to try several lines before you get one that fits
well, or maybe none of them will fit well). Insert a screenshot of the line that fits best
below.
e. Click on the line you added to your scatterplot. What is the equation for
that line? Write it below.

y= -44.5134+0.131654 x

f. How well do you think the line fits your data?

Very Well

g. Now, you can have your computer help you decide how the line fits your
data. Insert a calculator page. Go to Tools > Statistics > Stat Calculations > Two-variable
Statistics. Click ok, then select the two variables you are using for X List and Y List. Click
ok. What is the number that goes with r

r = 0.952499

h. Use the table to the right to find out what sort Weak vs Strong and Positive vs
of association your variables have. Write your Negative Association/Correlation -
answer below: USING r - VALUE

-1 to -0.7 = strong negative


0.95, strong positive. -0.69 to -0.3 = weak negative
-0.29 to 0.29 = none
0.3 to 0.69 = weak positive
0.7 to 1 = strong positive

The example above has a correlation


coefficient of 0.73, so kneeling height
and height have a strong positive
correlation.
i. Choose a point on your scatterplot and delete it from your spreadsheet. What is the new
equation?

y=44.1956+0.131286

j. How well do you think the line fits your data?

Very well

k. Now, you can have your computer help you decide how the line fits your data. Insert a
calculator page. Go to Tools > Statistics > Stat Calculations > Two-variable Statistics. Click ok,
then select the two variables you are using for X List and Y List. Click ok. What is the number
that goes with r

r = 0.948636

L. Use the table to the right to find out what sort Weak vs Strong and Positive vs
of association your variables have. Write your Negative Association/Correlation -
answer below: USING r - VALUE
r = 0.948636, so there is a strong positive
correlation. -1 to -0.7 = strong negative
-0.69 to -0.3 = weak negative
-0.29 to 0.29 = none
0.3 to 0.69 = weak positive
0.7 to 1 = strong positive

The example above has a correlation


coefficient of 0.73, so kneeling height
and height have a strong positive
correlation.

m. Did removing that point sigificantly affect your line of best fit, and your correllation coefficient?

No, the line of best fit and correlation coefficient stayed relatively the same.
IF YES, THEN THE POINTT YOU REMOVED WAS AN INFLUENTIAL POINT. INFLUENTIAL points look like
outliers, but greatly affect the data if removed.

n. Create a claim from your boxplot and histogram (Hint: Do you think these variables are related?).
Write it below.

The steady increase of registered boat usage has a strong positive correlation with how many
manatees killings occur each year.

o. What is your edivence? Why do you think your claim is true?

The scatterplot above shows that the more boats that are registered, the more manatees die.

Das könnte Ihnen auch gefallen