CHARTING BASICS The Garlic press
Excel a beginning! 2000 The Garlic Press
SPREADSHEET BASICS EXCEL BASICS EXCEL CHARTS EXCEL DATA BASE
BACK

Contents:

Introduction

Spreadsheet (Excel, Lotus)  and statistical programs (SPSS, SAS, STATA) are very powerful. They allow anyone who is willing to spend the time to generate useful numbers of just about everything.  Sometimes, however, to truly get the point across, a chart of the data can say what is needed much more efficiently than the data itself.  Charts can also be used to provide insights into relationships and distributional characteristics that the data or typical linear statistics fail to show. For both of these tasks the charting features of Excel, Clarisworks, Microsoft Works and  SPSS  are excellent tools. Personally I perfer Excel as my basic chart creating tool.

Charts or graphs have advantages over raw data in several different aspects.  They allow a visual presentation. To look at a well designed chart can sometimes provide possible correlations many times faster than examination of the data itself.  Why would one use charts over just a standard spreadsheet?

Why use Charting

To Explore:  Many times we would like to see if there is a relationship between variables.  Suppose that you wanted to determine if there is a relationship between: a countryâs GNP and the infant mortality rate, between age and the number of deviant acts one may have committed ("opportunity theory"), or between gender and one's view on abortion.  It may be quicker and easier to create a chart to immediately see the possible relationship of variables to one another, rather than paging through raw data or developing and interpreting statistics on the data.

To Present:  It seems that at this day and age, at least in American society, we want to be provided with as much information in as little time as possible.  Again, this is where graphing plays a key roll.  It seems that there is no longer any time to sit and read a newspaper in order to find out what is going on.  However, newspapers, such as USA Today  and weekly magazines such as Time and Newsweek (which were early users of charting techniques), seem to understand this phenomena and provide graphs to convey and sum up ideas that they are making in their articles.  Presenting information in graphs provides different ideas in smaller amounts of time. As the Chinese proverb says, ãa picture is worth a thousand words.ä

To Convince:  The same way that a graph can be used to present and explore different characteristics of data, it can also be used to convince.  Graphs have the ability to take large amounts of information and make them into exhibitions that are easily used to persuade.  And, as we will comment on, graphs also provide an easy way to distort data to make the point you want to make, or in the common student vernacular, "lie with statistics".

A Sample Spreadsheet & Chart

Our Data:
Our Chart
(this chart of part of our data shows the major components of a chart--it is not a good chart--too cluttered,
but more on that later):
 
Charting Terms
A. Title
B. Legend
C. Category, X axis, 
D. Data series, Y axis, scale
E. Data point
F. Gridlines
G. Unattached text
             (footnote reference, description, etc.)
Types of Charts
Excel offers in its gallery of choices eight chart types plus combinations, and 3-D versions of the eight types. Additionally each chart type will offer a gallery selection of autoformats for the chart type chosen. These are sets of chart formatting rules, such as use or don't use legends, use or don't use grid lineslines, etc. At first this may seem a tad overwhelming,but, with a little practice and careful consideration of what you want your chart to accomplish and what is appropriate, much of this confusion can be eliminated. Best of all, any choice you make is not final, you can easily change or modify your chart choice.

Before creating your presentation chart, take time to analyze and to decide on what chart you want, and then prepare your spreadsheet. Most important is to remember that the major purpose for most charts is to provide clarity and save time for the reader. In your choice, consider the following possible charts.

Column Charts:
Very good for comparing quantities of 2 or more groups for a variable over a particular period of time (e.g. comparing educational level, income, life expectancy, etc. for different ethnic groups in a population).Column charts are also good for showing how the quantity of a single variable changes over a particular period of time (ex. women's earnings as a percentage of men's for each year 1984-1994). In the latter case you want to emphasize the quantities themselves. See "Hints" under Bar Charts.

Bar Charts
These generally are used to display the same data as column charts except that each category is a row instead of a column. You can have a lot more labeled categories in a bar chart since you have the length of the page for different categories while a column chart has only the page width. "Histograms" are bar charts where each bar represents combined frequencies of several categories rather then the typical 1 category = 1 bar ( e.g. social science articles frequently group an age variable into a separate bar for each 10 year period 1-10, 11-20, 21-30....)

Hints: If you have too many bars or columns, group them in some logical or theoretical grouping, and then make a summary chart of the group. If you have lots of data points consider a line chart

Stacked Column or Bar Chart
These are bar or column charts with each bar/column broken down into components of the whole bar/column. For example if we had a
chart representing CSUB enrollment with each bar being a quarter (fall, winter, spring) and each bar broken into first quarter freshmen, first quarter transfers and returning students. See "Hints" under Bar Charts
 
 

TO BE CONTINUED