home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!gatech!paladin.american.edu!auvm!AKRONVM.BITNET!R1RLE
- Message-ID: <STAT-L%92090315435921@VM1.MCGILL.CA>
- Newsgroups: bit.listserv.stat-l
- Date: Thu, 3 Sep 1992 14:48:54 EDT
- Sender: "STATISTICAL CONSULTING" <STAT-L@MCGILL1.BITNET>
- From: Rich Einsporn <R1RLE@AKRONVM.BITNET>
- Subject: color analysis
- Lines: 33
-
- Researchers from Home Economics have asked me for statistical help with a study
- they are doing on color analysts. Two female subjects each visited the same 10
- color analysts. Each color analyst classified each subject according to what
- they call "season" -- spring, summer, autumn, or winter -- based on their
- perception of the subject's skin, hair, and eye coloring. The analysts also
- gave swatches of material of different color shades that they recommended for
- the subject.
- First off, the researchers want to infer something about whether the
- analysts are in agreement as to which season a subject belongs. Is there a
- standard measure out there to assess the degree to which the 10 analysts agree
- on the "season" of a given subject? Can that measure be "pooled" somehow
- to encompass both subjects at once?
- Second, the researchers have spent untold hours comparing the color swatches
- that were given to the subjects by the color analysts. First, they classified
- each swatch as belonging to one of about 13 basic hue categories. Then, within
- each hue category they compared the swatches given to one of the subjects by 2
- of the 10 analysts. (This was done for all 10 choose 2 pairs of analysts for
- each of the two subjects!) For each of these hue group swatch comparisons they
- counted the number of matching and nonmatching swatches provided by the 2
- analysts. (Recall that they're comparing 2 analysts at a time.) To add to the
- confusion, the number of swatches in a given hue group differed from analyst
- to analyst. Can anybody out there suggest anything reasonable to do with this
- information in terms of assessing the degree to which the color analysts were
- in agreement with their swatch selections?
- If two analysts agree on the subject's "season" then their swatch selections
- should be more similar than for two analysts whose "season" determinations
- differed. But I can't even think of a crude way testing this.
-
- Any suggestions would be appreciated.
-
- Dr. Rich Einsporn R|RLE@AKRONVM
- Dept. of Mathematical Sciences
- The Univ. of Akron
-