The following bivariate data set contains an outlier.

Question

The following bivariate data set contains an outlier.

xy
80.9	-72
90.8	-498.6
81.6	1519.7
60	-179.3
81.8	341.3
78.2	-1052.8
74.2	2241.4
82.9	-8.7
71.2	-382
85.2	-456.5
75.1	803.2
80.5	-1011.9
88	144.8
68.3	157.8
183.2	-828.6

What is the correlation coefficient with the outlier?

[Round your answer to three decimal places.]

r_w = 

What is the correlation coefficient without the outlier?

[Round your answer to three decimal places.]

Answer 1

Hi Emma,

Take either a graphing calculator (i.e. TI-80s series) or a statistical software program. In TI-83 Plus, make two lists--L1 and L2. Enter the data above as-is. Then, go to STAT-CALC-LinReg ax + b--enter L1, L2. You will get the regression line equation, r² (aka coefficient of determination), and the correlation coefficient r, which is what the question asked for.

I got:

r_w = 0.083 for data with outlier included

Now, to identify the outlier, go to 2nd-STAT PLOT-enter L1, L2. You may need to adjust the window, which you can do with your window key right next to STAT PLOT. The outlier will be apparent once you do this, so I will leave that determination to you. Once you identify the outlier, take both its x and y coordinate off the lists and repeat the above procedure.

I hope this helps.