Login

Welcome, Guest. Please login or register.

November 01, 2025, 10:11:26 am

Author Topic: Parallel box plots.  (Read 1021 times)  Share 

0 Members and 1 Guest are viewing this topic.

cindyy

  • Victorian
  • Forum Obsessive
  • ***
  • Posts: 254
  • RAH RAH AH AH AH ROMA ROMAMA GAGA OOH LA LA
  • Respect: +1
Parallel box plots.
« on: February 28, 2010, 10:01:23 pm »
0
What are some ways of deducing the presence of a relationship in a parallel box plot?

I know you look at the median, IQR, Rang and shapes, but the essential book doesnt explain how they should look like in order for a relationship to be present.
2009: Chinese(SL), Psychology
2010: Biology, Chemistry, English Language, Further, Methods CAS


Gloamglozer

  • The Walking VTAC Guide
  • Honorary Moderator
  • ATAR Notes Legend
  • *******
  • Posts: 4170
  • Here to listen and help
  • Respect: +324
Re: Parallel box plots.
« Reply #2 on: March 02, 2010, 03:56:21 pm »
0
What are some ways of deducing the presence of a relationship in a parallel box plot?

I know you look at the median, IQR, Rang and shapes, but the essential book doesnt explain how they should look like in order for a relationship to be present.

With boxplots, there's:

1.  The distribution (page 46 in the non TIN/CP version of Essentials)
2.  Whether or not it has outliers.
3.  The five-number summary (Minimum, Q1, Median, Q3 & Maximum). 
4.  Then there's the spread, as measured by both the range and the IQR. 

That's all you need to include in your "report" when comparing two boxplots.

Is that what you're after?

Bachelor of Science (Mathematics & Statistics) - Discrete Mathematics & Operations Research

crayolé

  • Victorian
  • Forum Leader
  • ****
  • Posts: 788
  • Respect: +3
Re: Parallel box plots.
« Reply #3 on: March 02, 2010, 05:31:16 pm »
0
Basically, If theres a change in outliers, median, range or IQR there is a said relationship between the variables.