box plots - easy question

mooshoepork

Trying to do a box plot and comment on the skewness for an assignment I have to do. Fairly easy stuff right? wrong

Excel is giving me a skewness of 0.17, which I'd assume is a slight positive skew

However, my calculator's boxplot is definiately shows that it's a negative skew?

here is my data

45
49
56
57
55
40
59
44
50
59
54
33
47
60
35
53
41
45
30
55
52
34
60
30
43
51
48
71
75
47
72
55

any ideas? I don't want to fuck it up. It's an easy assignment, it'd be great if I could get 100%. Just not sure what they want me to write when I'm getting conflicting answers.

They also said to use IQR to draw it, which it doesn't look like you even need to (it's been a while I'll admit), I just found the max, min, q1,q3 and median and drew it up.

Daenris

Why do you think it's a negative skew based on the calculator plot?

The number from Excel is right (or approximately right), and it's a very slight positive skew.

What are the numbers you're using to plot for min/max/q1/q3/median? Since it is such a small skew, if you have one or two of these wrong, the boxplot could look very different.

mooshoepork

Well the median is slightly on the right side, which would lead me to beleive it is negative. Like

---

[ | ]

---

Like that.


straight from excel

median 50.5
min 30
max 75
q1 43.5
q3 56.5
skew 0.17

Daenris

The median is slightly on the left side, not right. The min is 30, the median is around 50, and the max is 75. 30 to 50 is 20, but 50 to 75 is 25, so the median is slightly to the left, leaving a larger tail of numbers to the right, for a slight positive skew.

Edit: Unless you mean the median is slightly to the right of the actual box in the boxplot. This isn't what skewness is testing though. Skewness basically just tells you which side of the plot has a longer tail, and in this case it's the right side, so a positive skew.