MAT 240: Excel Project 3 Name:

As previously noted in Excel Project 1, one of the most popular types of psychological tests in use today
are intelligence tests. A long-standing question that psychologists have attempted to answer over the years
is whether IQ scores are mainly the result of people’s inherited traits (“nature”) or mainly the result of
their environment (“nurture”). Inherited traits include genes and hereditary factors while environmental
factors include a person’s upbringing with respect to childhood experiences, social relationships, and the
surrounding culture.

In 1981, clinical child psychologist Susan Farber 1 conducted a cross-sectional study to measure the effect
of home environment on intelligence. In order to research this question, Farber conducted a “twin study”
to compare IQ scores for identical twins, one of whom was raised by birth parents and the other by
someone else. Utilizing identical twins allowed Farber to control for possible genetic differences that may
affect intelligence (genetic differences would be considered a confounding factor). The design Farber
utilized is a matched pairs design. Data are considered paired (or dependent) when the observations in
one sample are matched directly to the observations in the second sample. The observations in Farber’s
study were paired by twin. This design allowed Farber to compare individuals who differed with respect
to the explanatory variable (upbringing) but were considered as alike as they could possibly be with
respect to the response variable (inborn intelligence as measured by IQ scores).

In Excel Project 3, we are going to use data collected by Farber to investigate the research question
below.

Are there statistically significant differences in IQ scores between people who were raised by their
birth parents, and those who were raised by someone else?

The appropriate hypothesis test to use when testing for differences between two matched (dependent)
variables is a paired t-test. The steps for carrying out a paired t-test are covered in Section 11.2 of your
statistics textbook. In a paired t-test, the data from two samples are reduced to one sample of differences.
The hypotheses are stated in terms of the mean of the paired differences, where μd denotes the
presumed population mean of the paired differences.

For our research question, the null and alternative hypotheses are stated below. We will test at the α=0.05
level of significance.

H0: μd =0 (interpreted to mean that the population mean of the paired differences in IQ Scores (Birth
Parents – Other) is 0)
HA: μd ≠ 0 (interpreted to mean that the population mean of the paired differences in IQ Scores (Birth
Parents – Other) is a value other than 0)

Once the null and alternative hypotheses have been stated, the next step is to check that the requirements
for carrying out a paired t-test are met. Since the probability distribution that we use for testing means is
Student’s t-distribution, the requirements for carrying out a paired t-test are the same as the requirements
for carrying out a one sample t-test.

 The first requirement is that a simple random sample is used. While the samples of twins are not
random samples from the two populations, they can be considered random, assuming that there is
nothing special about the IQ of a person just because he/she has an identical twin.
 The second requirement is that the population is normally distributed or the sample size is at least
30. There are 32 pairs of data, so n = 32 which meets the second requirement.

1
Farber, S. (1981). Identical Twins Reared Apart: A Reanalysis. Basic Books.
The directions below will guide you through the procedure for carrying out a paired t-test as you use
Excel to analyze the twin data.

Procedure:

The first step in carrying out a paired t-test is to calculate the difference between the two observations on
each pair.

 Open the Excel workbook on Moodle titled “Matched Pairs Twin Data”. There are 32 data values
in columns A and B. Each row of data represents the IQ scores for one pair of twins. The data
values in Column A (Birth) are the IQ scores for the twin raised by birth parents. The data values
in Column B (Other) are the IQ scores for the twin raised by someone else.
 Label the first cell in Column C as “difference”
 To compute the difference between the two IQ scores for each pair of twins, click on the second
cell in Column C and type the following formula: =A2 – B2 and click Enter. You should see the
difference between the IQ scores for the first pair of twins (113 – 109 = 4). Use your cursor to
hover over the bottom right corner of cell C2. When you see a small black cross-hairs on the
corner of the cell, click and hold while dragging down Column C for the remaining pairs of data.
You should now see the differences between IQ scores for all of the twin pairs.

The next step in carrying out a paired t-test is to compute the test statistic and p-value. These values are
obtained by carrying out a one sample t-test on the paired difference data. Recall that the test statistic
formula for a one sample t-test is:

x́−μ
t=
s
( )
√n
where x́ is the sample mean, μ is the presumed value of the population mean from the null hypothesis, s
is the sample standard deviation, and n is the sample size. When using the differences between paired
data as the sample, we modify the notation in the test statistic formula as follows:

d́−μd
t=
sd
(√ )
n

where d́ is the sample mean of the paired differences, μd is the presumed value of the population mean
from the null hypothesis, sd is the sample standard deviation of the paired differences, and n is the sample
size. Since the presumed value of the population mean from the null hypothesis is 0 ( μd = 0), the test
statistic formula for a paired t-test simplifies to:

d́
t=
sd
( )
√n
We are now ready to use Excel to compute the test statistic and p-value.

 On the Excel worksheet, enter the labels in Column E as shown below.

 To find the sample mean of the paired differences, click on cell F2, click on the insert function
button fx beside the toolbar, select statistical as the category, select average as the function, and
click OK (see screenshot below). In the Functions Argument dialogue box (see screenshot
below), click inside the Number 1 box, click and drag the difference data in Column C, and click
OK. You do not have to insert anything in the Number 2 box. The sample mean of the paired
differences should appear in cell F2.
 To find the sample standard deviation of the paired differences, click inside cell F3, repeat the
same steps as above, except in the Insert Function dialogue box, select STDEV.S as the function
(see screenshot below). The sample standard deviation of the paired differences should appear in
cell F3.

 Enter the sample size in cell F4. There are 32 pairs of data, so the sample size is 32.
 To compute the test statistic, click on cell F6 and enter the following formula:
=F2/(F3/SQRT(F4)). Press Enter and the test statistic should appear in cell F6.
 To compute the p-value, click on cell F7, click on the insert function button fx beside the toolbar,
select statistical as the category, select T.DIST.2T as the function, and click OK (see screenshot
below). In the Function Arguments dialogue box (see screenshot below), type in the absolute
value of the test statistic for X. For example, if the test statistic is -2.345, type in 2.345. Excel
requires X > 0, and you’ll get an error if you type in a negative number. The degrees of freedom
for a paired t-test is the same as the degrees of freedom for a one sample t-test which is n – 1.
Type in the degrees of freedom in the Deg_freedom box and press Enter. The p-value should
appear in cell F7.
The next step in carrying out the hypothesis test is to draw a conclusion.

Method 1: Drawing a conclusion using Rejection Regions

Notice from the inequality symbol found in the alternative hypothesis (H A: μd ≠ 0), we are carrying out a
two-tailed test. Using Table C (Critical values of t), identify the critical t-values for a two-tailed test with
df = n – 1 and α = 0.05. Record your answers below.

Negative critical t-value = Positive critical t-value =

Recall that the decision rule for the rejection regions method is that we reject the null hypothesis if the
calculated test statistic falls within the rejection region. Record the value of the test statistic from your
Excel output below.

Test statistic =

To determine if the test statistic falls within the rejection region, the test statistic is compared to the
critical value(s). Based on a comparison of the values recorded above, circle the correct conclusion.

Reject the null hypothesis Fail to reject the null hypothesis

In the space below, provide a brief rationale for your conclusion.

Method 2: Drawing a conclusion using p-values

Recall that the decision rule for the p-value method is that we reject the null hypothesis if the p-value is
less than or equal to the level of significance α. Record the p-value from your Excel output below and the
level of significance α.

p-value = α=

Based on a comparison of the values recorded above, circle the correct conclusion.

Reject the null hypothesis Fail to reject the null hypothesis

The last step in carrying out the hypothesis test is to interpret your decision.

In the space below, interpret your decision in relation to the research question “Are there
statistically significant differences in IQ scores between people who were raised by their birth parents,
and those who were raised by someone else?”

In the space below, comment on other factors that may have confounded the results of Farber’s study.

Submit your Excel output and your recorded answers on Moodle per your instructor’s directions.

Important Note: In analyzing data that involve two population means, it is very important to pay
attention to whether the paired t-test or the two-sample t-test is appropriate. Being aware of the study
design is extremely important. If we had not used a matched pairs design to analyze the IQ scores data,
and had instead analyzed the data as if the two samples were independent using a two-sample t-test, we
would have obtained a p-value of 0.213. Using the wrong statistical method could lead you to wrong
conclusions, which could have serious implications.