Discussion: View Thread

Hello everybody!

I'd like to see how a characteristic (categorical variable with 5 values) is distributed into 3 countries. Is there any other way to go over it apart from frequency table or cross tabulation?

Thank you in advance,

Lina
------------------------------
Efthalia Massou
PhD candidate - Researcher
Panteion University of Social and Political Sciences
------------------------------

Hello everybody!

I'd like to see how a characteristic (categorical variable with 5 values) is distributed into 3 countries. Is there any other way to go over it apart from frequency table or cross tabulation?

Thank you in advance,

Lina
------------------------------
Efthalia Massou
PhD candidate - Researcher
Panteion University of Social and Political Sciences
------------------------------

I'm a big fan of stacked barcharts for what you want to do. A stacked barchart would be a direct representation of your 5x3 table, much more so than a regular barchart. In Excel, you can arrange it so that the stacked bars for your three countries are side by side, with each country having 5 categories stacked on top of each other so that they visually sum to 100%, and with each category in the stack showing its cell frequency as its label. You can copy-paste your 5x3 table of cell counts directly into Excel, use the Chart Wizard to select Stacked Barcharts, have the basics done in 2 or 3 minutes, and have it looking half-way pretty in 10-to-15 minutes.

Are your five categories nominal or ordinal?

------------------------------
Eric Siegel
Biostatistician
Univ of Arkansas for Medical Sciences of Biostatistics
------------------------------

A mosaic plot is another way to visualize two-way cross-tabulated data.

http://www.cyclismo.org/tutorial/R/intermediatePlotting.html

Here's some discussion of the merits of mosaic plots vs other

forms of visualization, including barcharts or histograms.

https://www.perceptualedge.com/articles/visual_business_intelligence/are_mosaic_plots_worthwhile.pdf

http://www.perceptualedge.com/example13.php

Mosaic plots can be useful, but it's certainly true that they have a learning curve, and some limitations. (Personally, I find them most useful when there's a natural order in the values of one of the variables, with a coloring scheme applied that aligns with that ordering--e.g., red to white to green for bad/neutral/good or low/medium/high, etc.)

>>Kathy

------------------------------
Katherine Godfrey
------------------------------

A mosaic plot is another way to visualize two-way cross-tabulated data.

http://www.cyclismo.org/tutorial/R/intermediatePlotting.html

Here's some discussion of the merits of mosaic plots vs other

forms of visualization, including barcharts or histograms.

https://www.perceptualedge.com/articles/visual_business_intelligence/are_mosaic_plots_worthwhile.pdf

http://www.perceptualedge.com/example13.php

Mosaic plots can be useful, but it's certainly true that they have a learning curve, and some limitations. (Personally, I find them most useful when there's a natural order in the values of one of the variables, with a coloring scheme applied that aligns with that ordering--e.g., red to white to green for bad/neutral/good or low/medium/high, etc.)

>>Kathy

------------------------------
Katherine Godfrey
------------------------------