Navigational Menu OVERVIEW OF STATISTICAL THINKING MICROCASE MICROCASE
|
MICROCASE BASIC STATISTICS UNIVARIATE STATISTICS Statistics
to describe one variable 1.
When you select the UNIVARIATE STATISTICS option, a window will
appear for entering the primary variable you are interested in--your
dependent variable. Type
in the variable # or type in the variable name (be careful about
spaces and periods). 2.
If you are not sure which variable you want, there is a search
option or you can scroll through the entire variable list.
If you highlight a variable in the variable list, the
definition of the variable (often the question asked of survey
respondents) and the possible attributes will appear in a box in the
lower right of the window. 3.
Sometimes you will want to look at a variable for only part of
the data set, for example, only for males or only for females, or only
for people with college educations, or only for people living with a
romantic partner. To do
this type a variable # or name in the SUBSET box.
For example, if you want to examine a variable only for men,
you would type SEX in the subset box; the computer would then prompt
you whether you want to look only at men or only at women, and you
would select men. You can
select more than one subset variables, for example use sex and marital
status to select married women. 4.
Be careful to eliminate Adon=t knows@ and Ano
answers@
by using the SUBSET function (or by defining them as missing data on
the FILE SETTINGS screen as you open the data file) before you write
down the statistics or copy the pie charts and bar graphs. 5.
When you=ve indicated the variable you=re interested in and any subset variables, click on OK
in the top right corner of the window. 6.
Three kinds of statistics are available from the UNIVARIATE
STATISTICS option: Pie
Charts:
Pictorial depictions of the percentage distribution of the variable.
If 100% of the sample gave one response or shared the same
attribute, the whole pie would be one color; the size of the different
colored pie wedges represents the proportion of the sample who share
that attribute. Bar
Graphs:
When there are more than 10 attributes in a variable, MicroCase
automatically gives you a bar graph rather than a pie chart.
Here the heights of the bars represent the number (frequency)
of the sample units which share each attribute. Statistics:
By clicking on the Statistics option on the
left margin of the screen, you can get a frequency and percentage
distribution, as well as the summary statistics.
There are three averages which you can see from this screen: MODE:
most frequently occurring score (score with the highest frequency or
percentage). This average is useful with categorical data (e.g., 1=men,
2=women) where there is no logical order to the categories; that is,
although men are scored 1 and women are scored 2, it is nonsense to
think men are less than women. --Note, when there is more than one score in the median interval, MicroCase interpolates the exact position of the median between the upper and lower limits of the median interval. --Note, sometimes GSS lists categories that you think could be ordered if they were rearranged (e.g., 1=agree, 2=disagree, 3=can=t choose); you can use the COLLAPSE option on the DATA MANAGEMENT menu to recode these responses (e.g., 1=agree, 2=can=t choose, 3=disagree). MEAN:
arithmetic average of the scores (add up all the scores and divide by
the number of scores). This
average is useful when the response categories are ordered numerical
categories, e.g., age in years, education in years, income in $s, etc. VARIANCE:
squared average variation around the mean. The variance is derived
by subtracting each individual score from the mean of the distribution
of scores (deviation scores), squaring each deviation score,
summing the squared deviation scores (sum of squares), and
dividing the sum by the number of scores in the
distribution. The mean for this distribution of scores is 9 ([8+9+10+9]/4). Alex's deviation score is -1, Jessie's deviation score is 0, Les' deviation score is +1, and Sam's deviation score is 0. The sum of the deviation scores is zero. This will be true for every distribution, assuming the mean and the deviation scores are calculated correctly. The intuitive mean deviation would be the sum of the deviation scores divided by the number of deviation scores, but this number will always equal zero. Therefore, statisticians square each of the deviation scores so that the sum of the deviation scores will not be zero unless every score is the same as the mean score. The average of the sum of the squared deviations (2/4=.5 in the example above) is called the variance. STANDARD
DEVIATION: the square root of the variance. The standard
deviation converts the variance back to raw score units (instead of
squared raw score units) and is interpreted as the average variation around the mean.
In the example above the square root of 0.5 is 0.71. A researcher
would describe the above distribution as having a mean of 9 and a
standard deviation of .7. In other words, the central tendency of
the distribution is 9 and on, average scores deviation from that central
tendency by 7/10ths of a point. CONFIDENCE
INTERVALS OF THE MEAN: the range of scores that a researcher is
confident--at different degrees of certainty--contain the actual
population mean.
STANDARD
SCORES (or z-scores): deviations above or below the mean in units of
standard deviation. 7.
You can print the pie chart, the bar graph, or the statistics by
clicking on the printer icon in the top tool bar. 8.
To copy a screen into a word processing document, click on the
double-sheet-of-paper icon in the top tool bar; this copies the screen
to the clipboard. Minimize
the MicroCase screen by clicking on the dash line in the very top right
toolbar. Open (or maximize)
your word processor (either Word or Word Perfect) and paste the screen
image into your word processing document.
The paste icon is a clipboard with a sheet of paper on top.
To return to MicroCase when you are finish with the word
processor, click on the MicroCase icon in the taskbar along the bottom
of your screen. 9.
Note that the MicroCase tool bar at the top also allows you to
view the FILE NOTES screen for the data set (by clicking on the notebook
icon) and the variable definitions (by clicking on the V icon).
If you open either of these documentation screens, be sure to
click on OK to close them. 10.
To look at the UNIVARIATE STATISTICS for another variable, click
on the counter clockwise arrow in the top tool bar. 11.
To move to another statistical option (e.g., CROSS TABULATION or
COLLAPSE) or to exit from MicroCase, click on the MENU button in the top
tool bar. |
|
for questions or comments
contact me at mduncombe@coloradocollege.edu |