EXERCISE ON t-test

INSTRUCTIONS: Solve the problems using manual calculation first. Repeat using Excel and
finally use SPSS. Submit your solutions in hard copy form to my pidgeon hole by the 29th
of November 2017.

1. A research study was conducted to examine the differences between older and younger
adults on perceived life satisfaction. A pilot study was conducted to examine this
hypothesis. Ten older adults (over the age of 70) and ten younger adults (between 20
and 30) were given a life satisfaction test (known to have high reliability and validity).
Scores on the measure range from 0 to 60 with high scores indicative of high life
satisfaction; low scores indicative of low life satisfaction. The data are presented below.
Is there a difference between older and younger adults on perceived life satisfaction?

Older Adults Younger Adults

45 34
38 22
52 15
48 27
25 37
39 41
51 24
46 19
55 26
46 36

a. What is the most appropriate t-test to use and why?

Independent t-test, because;

i. there are two group of data samples,

ii. two independent sample means

b. What would be the null hypothesis in this study?

Ho: there is no statistically significant difference on life satisfaction between
older and younger adults

c. What would be the alternate hypothesis?

Ha: there is a statistically significant difference on life satisfaction between
older and younger adults

1|Page
 d. What probability level did you choose and why?
.05 significance level or 90% confidence level (0.1/2), accepting that this has a greater
degree of uncertainty than 95% or 99%.

e. What is your tcrit?

df = n1 + n2 -2
= 18
= 2.101 tcrit

f. Is there a significant difference between the two means?

Older Adults, x Younger Adults, y x-x bar (x-x bar)2 y-y bar (y-y bar)2
45 34 0.5 0.25 5.9 34.81
38 22 -6.5 42.25 -6.1 37.21
52 15 7.5 56.25 -13.1 171.61
48 27 3.5 12.25 -1.1 1.21
25 37 -19.5 380.25 8.9 79.21
39 41 -5.5 30.25 12.9 166.41
51 24 6.5 42.25 -4.1 16.81
46 19 1.5 2.25 -9.1 82.81
55 26 10.5 110.25 -2.1 4.41
46 36 1.5 2.25 7.9 62.41
445 281 678.5 656.9
x bar = 44.5 y bar = 28.1

Sx = √∑ (x-x bar)2 Sp2 = (nx-1)Sx2 + (ny-1) Sy2

n-1 n1 + n2– 2
=√678.5 = 9 (75.34) + 9 (72.93)
9 18
= 8.68 = 36.47
Sx2 = 75.34
t= 44.5 – 28.1
Sy = √∑ (y-y bar)2 √75.34 + 72.93
n-1 10 10
= √656.9
= 16.4
9
3.85
= 8.54
Sy2 == 72.93 = 4.259 (t v)
Sy x 2 > sx, means pooled variance

2|Page
 2.101 4.259

Reject Ho
- There is a statistically significant difference on life satisfaction between
older and younger adults

g. Interpret your answer and make a conclusion

Older adults have significantly higher life satisfaction of younger adults ( t =
4.259, p<.01).

2. A teacher claims that his Science Module will increase students’ science knowledge by
an average of 31 points. You randomly select 8 students from your school and let them
use the Module for one semester and test their science knowledge. You get the
following science knowledge scores:

Science knowledge: 30 28 32 26 33 25 28 30

Does the actual science knowledge for these students deviate significantly from 31
(alpha = .05)?

a. What is the most appropriate t-test to use and why?

1 sample t-test, because we compare one group’s average value to a single
number

b. What would be the null hypothesis in this study?

Ho: µ < 31

c. What would be the alternate hypothesis?

Ho: µ >31

d. What probability level did you choose and why?

.05 confidence level, to avoid a type II error

3|Page
 e. What is your tcrit?
df = n-1
=7
tcrit = 1.895

f. Is there a significant difference between the two means?

x x-x bar (x-x bar)2

30 1 1
28 -1 1
32 3 9
26 -3 9
33 4 16
25 -4 16 -2.037 1.895
28 -1 1
30 1 1
232 54
x bar = 29

S = √∑ (x-x bar)2
n-1
= √54/7
= √7.71
= 2.777

t = x bar - µ
S/√n

= 29-31
2.777/√8

= -2
2.777/2.828

= -2
2.777/2.828

= -2
0.982

= -2.037

4|Page
 Reject H1, accept Ho: µ < 31
Science Module will not increase students’ science knowledge by an average
of 31 points

g. Interpret your answer and make a conclusion

Science Module will not increase students’ science knowledge by an average
of 31 points. (t = -2.037, p > .01)

3. Researchers want to examine the effect of perceived control on health complaints of

geriatric patients in a long-term care facility. Thirty patients are randomly selected to
participate in the study. Half are given a plant to care for and half are given a plant but
the care is conducted by the staff. Number of health complaints are recorded for each
patient over the following seven days. Is there a difference between health complaints
between the two groups? Solve the problem using manual calculation first. Repeat
using Excel and finally use SPSS

Control over Plant, x No Control over Plant, y x-x bar y-y bar (x-x bar)2 (y-y bar)2
23 35 6.4 7.94 40.96 63.04
12 21 -4.6 -6.06 21.16 36.72
6 26 -10.6 -1.06 112.36 1.12
15 24 -1.6 -3.06 2.56 9.36
18 17 1.4 -10.06 1.96 101.2
5 23 -11.6 -4.06 134.56 16.48
21 37 4.4 9.94 19.36 98.8
18 22 1.4 -5.06 1.96 25.6
34 16 17.4 -11.06 302.76 122.32
10 38 -6.6 10.94 43.56 119.68
23 23 6.4 -4.06 40.96 16.48
14 41 -2.6 13.94 6.76 194.32
19 27 2.4 -0.06 5.76 0.0036
23 24 6.4 -3.06 40.96 9.36
8 32 -8.6 4.94 73.96 24.4
249 406 849.6 838.89
x bar = 16.6 y bar = 27.06

a) What is the most appropriate t-test to use and why?

5|Page
 Independent t-test, because there are two sets of data sample, control and
non-control

b) What would be the null hypothesis in this study?

Ho: There is no significant effect of perceived control on health complaints of
geriatric patients between the control and non-control groups.

c) What would be the alternate hypothesis?

Ha: There is a significant effect of perceived control on health complaints of
geriatric patients between the control and non-control groups.

d) What probability level did you choose and why?

.05 to avoid Type II error

e) What is your tcrit?

Sx = √∑ (x-x bar)2
n-1
=√849.6
14
= 7.79
Sx2 = 60.69

Sy = √∑ (y-y bar)2
n-1
=√838.89
14
= 7.74
Sy2 = 59.92

Sx X 2 > Sy, so variances are equal.

Tcrit = n+n-2 = 28

= 1.701

f) Is there a significant difference between the two testing conditions?

Sp2 = (nx-1)Sx2 + (ny-1) Sy2

n1 + n2– 2

= 14 (60.69) + 14 (59.92)
28

6|Page
 = 849.66 + 838.88
28
= 60. 31

t = 16.6 – 27.06
√60.69 + 59.92
15 15

t = -3.69 1.701 3.69

Reject Ho: There is a significant difference on health complaints of geriatric patients

between the control and non-control groups

g) Interpret your answer and make a conclusion

Patients with control over a plant had significantly fewer health complaints
than patients with non-control over the care of the plant (t = -3.691, p < .01)

4. Suppose you wish to test the effect of Prozac (a medication for depression) on the well-
being of depressed students, using a standardised "well-being scale" that sums Likert-
type items to obtain a score that could range from 0 to 20. Higher scores indicate
greater well-being (that is, Prozac is having a positive effect). Do students feel less
depressed after taking Prozac?

Students Mood before Mood after D D2

Prozac, x Prozac, y
1 3 5 -2 4
2 0 1 -1 1
3 6 5 1 1
4 7 7 0 0
5 4 10 -6 36
6 3 9 -6 36
7 2 7 -5 25
8 1 11 -10 100
9 4 8 -4 16
X bar = 3.33 y bar = 7 -33 219

7|Page
 a) What is the most appropriate t-test to use and why?
Paired sample t-test because there is only one group of sample with two
sets of data (pre and post-test)

b) What would be the null hypothesis in this study?

Ho: There is no significance effect of Prozac on the well-being of depressed
students

c) What would be the alternate hypothesis?

Ha: There is a significance effect of Prozac on the well-being of depressed
students

d) What probability level did you choose and why?

.05 to avoid Type II error

e) What is your tcrit?

Df: n-1
= 9-1
t crit @.05 =8
= 1.8595

f) Is there a significant difference between the two means?

T= ∑d
Reject Ho
N
∑d2 – ( ∑d)2
N
N (N-1)

= -33 1.859 3.145

9 5
√219 - (-33)2
9
----------------
9 (8)

= 3.667
√219 - 121
72

= 3.667
√1.361

8|Page
 = 3.145

There is a significance effect of Prozac on the well-being of depressed

students

h) Interpret your answer and make a conclusion

A significant increase in well-being occurred: t(9) = 3.145, p = .014 in the Prozac

treated group compared to the control group.

9|Page