LESSON #1: DESCRIPTIVE STATISTICS

Measures of Central Tendency

A measure of central tendency is one value that represents a whole set of scores.

Three measures of central tendency are the MEAN, the MEDIAN, and the MODE.

To calculate the MEAN in Excel, you open up a blank workbook. Then type in the values in your
set in a column (in this example, I am using 84, 93, 99, 100, and 97). In the row beneath the
column of numbers that you entered, type in =AVERAGE( and then highlight the column of cells
for the rows that you just entered, and add a closing parenthesis ) and then hit ENTER on your
keyboard.

NOTE: When you type the equals sign into Excel, it tells Excel that what follows the equals sign
is a formula and so Excel will calculate the answer to the formula for you.

Another way to calculate the mean is to type in (for example): =(84+93+99+100+97)/5 (the 5 is
the total number of values).

To calculate the MEDIAN in Excel, type in =MEDIAN( and then highlight the column of cells for
the values in your set, and add a closing parenthesis ) and then hit ENTER on your keyboard.
To calculate the MEDIAN by hand, you first put the values in order from smallest to largest:

84, 93, 97, 99, 100

After you put the values in order, you just take the middle value. That represents the MEDIAN.
If you have an even number of values (as opposed to the five values that we have here), then you
just take the average of the middle two values.

To calculate the MODE in Excel, type in =MODE( and then highlight the column of cells for the
values in your set, and add a closing parenthesis ) and then hit ENTER on your keyboard.

In this case, Excel doesn't give you a MODE because every value is represented once.

The mode is the most common value (the value/number that occurs most often). A data set can
have no MODE (if every value is included an equal number of times), unimodal (one mode),
bimodal (two modes), trimodal (three modes), or multimodal (four or more modes).

Which one do we use?

In psychology research, you pretty much always use the MEAN.

However, in certain cases, the MEDIAN is useful. An example is when you have an extreme
value (known as an OUTLIER). An outlier is a crazy value that lies far away from most or all of
the others.

For example, if we added 23 to our set of values and recalculate the MEAN and the MEDIAN, we
would get:

Our MEAN changed from 94.6 to 82.667, but our MEDIAN only changed from 97 to 95.

Why didn't the MEDIAN change much?

Because the 23 is an OUTLIER on the end and when we order the values (23, 84, 93, 97, 99,
100), we're now taking the average of 93 and 97.

The MEAN is more sensitive to OUTLIERS. When there are big OUTLIERS, the MEDIAN is
useful to report in addition to the MEAN because it's a better description of the center of the
data set than the MEAN is when there are OUTLIERS.

But what about the MODE? When do we use that?

The MODE is useful when you have categorical data (nominal or ordinal data). For example, you
can’t take the MEAN of a Democrat and a Republican.

Standard Deviation

Next, we assess the STANDARD DEVIATION. We measure how much the values vary, or how
spread-out they are.

To calculate the STANDARD DEVIATION in Excel, type in =STDEV.S( and then highlight the
column of cells for the values in your set, and add a closing parenthesis ) and then hit ENTER on
your keyboard.

What does that mean?

Let’s map out the STANDARD DEVIATION.

Column #1: We will call each value X.

Column #2: What is the deviation of X? It is each individual value minus the MEAN.

NOTE: The theory on STANDARD DEVIATION is that we want to get a measure of the range of
scores (the variation or how spread out they are). The RANGE is the highest value minus the
lowest value. However, by just calculating the RANGE, we're not getting a sense of how each
individual value varies.
If we add the deviations up, it's going to equal zero and that's not very useful for assessing the
variation.

Therefore, what we do is that we take (X - the mean) and then we square each one.

This is our SQUARED DEVIATION. Which is the foundation for standard deviation
calculations.

Then we add these up in Excel by typing in =SUM( and then highlight the column of cells for the
squared deviations, and add a closing parenthesis ) and then hit ENTER on your keyboard.

This sum is a measure of collective deviation. We call this sum the SUM OF SQUARES.

NOTE: We do this to get rid of the negatives and to make these deviation scores meaningful.
When we square something all the negatives turn to positives.

To get the VARIANCE, we divide the SUM OF SQUARES by n minus one (n-1).

n is the number of values in our sample (in this case, 5)

Therefore, we will divide by four.

To get the STANDARD DEVIATION, we find the square root of the variance. To calculate this in
Excel we type in =SQRT( and then highlight the VARIANCE and add a closing parenthesis ) and
then hit ENTER on your keyboard.

REMEMBER: VARIANCE, also known as the mean square (MS), equals STANDARD
DEVIATION squared.

To describe STANDARD DEVIATION in words, it is basically the average deviation of an

Why do we use n - 1?

To calculate the MEAN average, we add up all the values and divide by n. In calculating our
VARIANCE, we use n-1 because we're trying to get our SAMPLE to represent a POPULATION.

To do this we need to make an adjustment by using n-1 in the bottom part of the fraction.

This allows us to make our sample standard deviation bigger (more representative of the
population standard deviation).

This makes the standard deviation from the sample an unbiased estimate.

Remember, the bottom part of the fraction becomes smaller to make the standard deviation of
the sample look more like the standard deviation of population and that's because the standard
deviation of the population is generally bigger.

Introduction to SPSS

SPSS will give you many more options for hypothesis tests than Excel.

NOTE: When you first start SPSS up, you might have to cancel or close a pop up screen that
shows up at the beginning. When you close that, you'll get to the blank spreadsheet.

The key to using SPSS is to remember to go to VARIABLE VIEW first and name your variables.
VARIABLE VIEW is a tab at the bottom.

In this example, we will call our variable “Temperature.”

NOTE: The .00 after each value will be added in automatically. So type in 84 hit ENTER and it
will come out 84.00. Some students that are new to SPSS type in the .00. You don't have to do
that. Just type in the whole number and let SPSS add the decimal places.

To get our descriptive statistics, go to ANALYZE > DESCRIPTIVE STATISTICS >

Next, click the arrow to move the name of your variable over to the variables column.
Then, click on STATISTICS and click on MEAN, MEDIAN, MODE, STANDARD DEVIATION,
and VARIANCE.

Click CONTINUE and then click OK.

SPSS will open up a separate window for the output.

Please submit your homework as one Excel file (per homework assignment). For the SPSS
output charts, just copy and paste the ones that are needed to your Excel spreadsheet.

THE END
(for now)