Professional Documents
Culture Documents
6.5B..................38
Introductory Statistics
Using SPSS 2nd Edition
Knapp Solutions Manual
Full download at link:
https://testbankpack.com/p/solution-manual-
for-introductory-statistics-using-spss-2nd-
edition-knapp-1506341004-9781506341002/
Chapter 6
ANOVA and Kruskal-Wallis Test
Solutions to All Exercises
Exercise Page
6.1A....................2
6.1B....................6
6.2A..................10
6.2B..................14
6.3A..................18
6.3B..................22
6.4A..................26
6.4B..................30
6.5A..................34
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
2
Exercise Page
6.6A ................. 42
6.6B ................. 46
6.7A ................. 50
6.7B ................. 54
6.8A ................. 58
6.8B ................. 62
6.9A ................. 66
6.9B ................. 71
6.10A ............... 76
6.10B ............... 81
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
3
(a)
H0: Practicing meditation has no effect on resting pulse rate.
H1: Practicing meditation reduces resting pulse rate.
(b) Histograms with normal curve plots show a normal distribution of pulse for the
groups as shown in the three figures below, hence, the pretest criteria of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
4
Normal distribution for pulse in Group 2 (meditated 30 minutes a day, 3 days per week)
Normal distribution for pulse in Group 3 (meditated 30 minutes a day, 6 days per week)
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
5
The homogeneity of variance score shows a significance (p) of .920; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference among the variances of the three groups, hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below is 35 for each group;
since the ns are greater than 30, this criterion passes also.
(c)
The ANOVA revealed the following:
Descriptives
pulse
ANOVA
pulse
Multiple Comparisons
pulse
Tukey HSD
The Tukey post hoc test was used since the ns for each group were the same (35
each).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
6
NOTE: Since the ANOVA test renders results involving multiple comparisons, it
may be helpful to organize the findings as shown in the table below. SPSS does
not generate this table directly, but you can construct it manually. You can copy
the group names and means from the Descriptives table, and the p values from
the Sig. column in the Multiple Comparisons table.
(d)
This study analyzed the effects that meditation had on resting pulse rates. The subjects
were randomly assigned to one of three groups; members of the control group did not
meditate, those in the second group meditated for 30 minutes on Monday, Wednesday
and Friday, and members of the third group meditated for 30 minutes Monday through
Saturday. After 2 weeks, those who meditated (3x / wk., μ = 92.20; 6x / wk. μ = 91.40)
showed a statistically significant reduction (p = .007 and p = .001, respectively) in
resting pulse rate compared to those who did not meditate (μ = 97.40) using a .05
level. We found no statistically significant difference in the resting pulse rates between
those who meditated 3 days per week, compared to those who meditated 6 days per
week (p = .881).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
7
(a)
H0: Practicing meditation has no effect on resting pulse rate.
H1: Practicing meditation reduces resting pulse rate.
(b)
Histograms with normal curve plots show a normal distribution of pulse for the three
groups as shown in the three figures below, hence, the pretest criteria of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
8
Normal distribution for pulse in Group 2 (meditated 30 minutes a day, 3 days per week)
Normal distribution for pulse in Group 3 (meditated 30 minutes a day, 6 days per week)
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
9
The homogeneity of variance score shows a significance (p) of .786; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference among the variances of the three groups, hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below are 33, 35, and 31;
since the ns are greater than 30, this criterion passes also.
(c)
The ANOVA revealed the following:
Descriptives
pluse
ANOVA
pluse
Multiple Comparisons
pluse
Sidak
The Sidak post hoc test was used since the ns for each group were not the same (ns =
33, 35, 31).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
10
Per the table above, practicing meditation for 2 weeks produced no statistically
significant reduction in resting pulse rate using a .05 level. Based on these findings,
we would not reject H0, and we would reject H1.
(d)
This study analyzed the effects that meditation had on resting pulse rates. The subjects
were randomly assigned to one of three groups; members of the control group did not
meditate, those in the second group meditated for 30 minutes on Monday, Wednesday
and Friday, and members of the third group meditated for 30 minutes Monday through
Saturday. After 2 weeks, resting pulse rates were recorded for each participant. Those
who did not meditate had a mean resting pulse rate of 95.15, which was slightly higher
than those who meditated (3 days per week: μ = 93.54, 6 days per week: μ = 92.68);
however, we detected no statistically significant differences among any of the three
groups using a .05 level, suggesting that the meditation schedules tested were not
effective in reducing resting pulse rates.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
11
(a)
H0: Mentorship has no effect on freshman’s term grades.
H1: Mentorship enhances freshman’s term grades.
(b)
Histograms with normal curve plots show a normal distribution of term scores for all
groups as shown in the three figures below; hence, the pretest criterion of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
12
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
13
The homogeneity of variance score shows a significance (p) of .114; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below, is 25 for each group;
although the ANOVA becomes more robust with an n of at least 30 per group, we’ll
proceed with the results given.
(c)
The ANOVA test revealed the following:
Descriptives
grade
95% Confidence Interval
for Mean
Std. Std. Lower Upper
N Mean Deviation Error Bound Bound Minimum Maximum
No mentor 25 68.84 7.565 1.513 65.72 71.96 53 79
In-person mentor 25 69.12 7.322 1.464 66.10 72.14 57 85
E-mentor 25 75.28 10.616 2.123 70.90 79.66 50 90
Total 75 71.08 9.025 1.042 69.00 73.16 50 90
ANOVA
grade
Multiple Comparisons
grade
Tukey HSD
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
14
There was no statistically significant difference in GPA between those who had no
mentor, and those who had in-person mentoring. Students in the E-mentoring had
GPAs statistically significantly higher than the other two groups. These findings lead us
to reject H0 and not reject H1.
(d)
This study assessed the impact that providing a sophomore to provide mentorship to a
freshman has on freshman grades. Freshmen were randomly assigned to one of three
groups: Group 1 received no mentor, Group 2 received in-person mentorship, and
Group 3 received e-mentorship (used digital communication). There was no statistically
significant difference between the GPAs for students who had no mentoring (μ = 68.84)
compared to those assigned to an in-person mentor (μ = 69.12) (p = .993), however,
students assigned to the e-mentoring group (μ = 75.28) statistically significantly
outperformed the No-mentor group (p = .027) and the In-person mentor group (p = .036)
using an level of .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
15
(a)
H0: Mentorship has no effect on freshman’s term grades.
H1: Mentorship enhances freshman’s term grades.
(b)
Histograms with normal curve plots show a normal distribution of term scores for all
groups as shown in the three figures below; hence, the pretest criterion of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
16
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
17
The homogeneity of variance score shows a significance (p) of .914; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below, is 25 for each group;
although the ANOVA becomes more robust with an n of at least 30 per group, we’ll
proceed with the results given.
(c)
The ANOVA test revealed the following:
Descriptives
grade
ANOVA
grade
Multiple Comparisons
grade
Tukey HSD
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
18
The Tukey post hoc test was used since the ns for each group were the same (25
each).
The mean GPAs for each of the two mentoring groups statistically significantly
outperformed those in the control group, who received no mentoring. Also, even though
the e-mentoring group had a slightly higher term grade than the in-person mentoring
group, this was not a statistically significant difference. These findings lead us to reject
H0 and not reject H1.
(d)
This study assessed the impact that providing a sophomore to provide mentorship to a
freshman has on freshman grades. Freshmen were randomly assigned to one of three
groups: Group 1 received no mentor, Group 2 received in-person mentorship, and
Group 3 received e-mentorship (used digital communication). The GPA for those who
had in-person mentoring (μ = 69.40) statistically significantly outperformed those who
had no mentoring (μ = 63.56; p = .017, = .05); similarly, those who had e-mentoring (μ
= 69.88) statistically significantly outperformed the no-mentoring control group (p =
.009). Although those in the e-mentor group earned a slightly higher GPA than those in
the in-person mentoring group, the difference is not statistically significant (p = .971),
hence, either form of mentoring seems to help enhance freshman GPAs.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
19
(a)
H0: Tending to a plant has no effect on depressive mood.
H1: Tending to a plant reduces depressive mood.
(b)
Despite the few low score outliers shown in the histogram for Group 2, the Histograms
with normal curve plots exhibit a normal distribution of the depress variable for all
groups as shown in the three figures below, hence, the pretest criteria of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
20
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
21
The homogeneity of variance score for mood shows a significance (p) of .133; since this
is greater than the level of .05, this suggests that there is no statistically significant
difference between the variances among the three groups, hence, this pretest criterion
passes.
The n for each group is 60 (see Descriptives table below), which satisfies the 30 per
group minimum criterion.
(c)
The ANOVA test revealed the following:
Descriptives
depress
ANOVA
depress
Multiple Comparisons
depress
Tukey HSD
Groups (μ = depression) p
No plant (μ = 19.60) : Bamboo (μ = 18.05) *.003
No plant (μ = 19.60) : Cactus (μ = 19.63) .997
Bamboo (μ = 18.05) : Cactus (μ = 19.63) *.003
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
22
There is no statistically significant difference in the average scores between those who
received no plant (μ = 19.60) and those who received a cactus (μ = 19.63) (p = .997,
= .05). The mean depression score for those who received a bamboo plant (μ = 18.05)
was statistically significant lower than those who received no plant (p = .003) and those
who received a cactus (p = .003).
Since those in the Bamboo group had a statistically significantly lower depression score
than those who received no plant, we would reject H0. By that same reasoning, we
would not reject H1.
(d)
We hypothesized that empowering nursing home residents with an opportunity to
provide nurturance would help reduce depression. To test this hypothesis, 180 residents
were randomly assigned to one of three groups: Those in Group 1 constituted the
control group, and were given no plant. Those in Group 2 were given a small bamboo
plant to tend to along with a card providing care instructions. Participants in Group 3
were given a cactus plant along with a card providing care instructions. After 90 days,
we administered the Acme Depression Scale (1 = Low depression, 100 = High
depression) to members of all three groups. We found that those who were given the
bamboo plant scored an average of 18.05; using a .05 level, we found that their
depression level was statistically significantly lower than those who were given no plant
(μ = 19.60, p = .003), and those who were given a cactus (μ = 19.63, p = .003). We
found that those who received a cactus had a slightly higher average depression level
(μ = 19.63) than those who were given no plant (μ = 19.60), however, there was no
statistically significant difference in depression scores when comparing those who were
given no plant to those who were given a cactus (p = .997). These findings suggest that
tending to a small plant has the potential to reduce depression in nursing home
residents, but the cactus, which essentially requires no tending, did not provide the
desired effect, whereas the bamboo, which required monitoring and watering, did.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
23
(a)
H0: Tending to a plant has no effect on depressive mood.
H1: Tending to a plant reduces depressive mood.
(b)
The Histograms with normal curve plot exhibits a normal distribution of the depress
variable for all groups as shown in the three figures below, hence, the pretest criteria of
normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
24
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
25
The homogeneity of variance score for mood shows a significance (p) of .333; since this
is greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of among the three groups, hence, this pretest
criterion passes.
The ns for the groups are 58, 51, and 65 (see Descriptives table below), which satisfies
the 30 per group minimum criterion.
(c)
The ANOVA test revealed the following:
Descriptives
depress
ANOVA
depress
Multiple Comparisons
depress
Sidak
Groups (μ = depression) p
No plant (μ = 15.69) : Bamboo (μ = 16.37) .760
No plant (μ = 15.69) : Cactus (μ = 17.03) .188
Bamboo (μ = 16.37) : Cactus (μ = 17.03) .766
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
26
Inspection of the Sig. (p) figure in the ANOVA table (p = .186, which is greater than the
.05 level) tells us that there is no statistically significant difference(s) detected in the
depression scores among any of the groups.
Upon reviewing the comparisons presented in the Multiple Comparisons table, we see
that this finding is confirmed; we see that the p level for each pair of depression scores
is greater than the specified .05 level, hence there are no statistically significant
differences between any of the groups as shown in the table above.
(d)
We hypothesized that empowering nursing home residents with an opportunity to
provide nurturance would help reduce depression. To test this hypothesis, 174 residents
were randomly assigned to one of three groups: Those in Group 1 constituted the
control group and were given no plant. Those in Group 2 were given a small bamboo
plant to tend to along with a card providing care instructions. Participants in Group 3
were given a cactus plant along with a card providing care instructions. After 90 days,
we administered the Acme Depression Scale (1 = Low depression, 100 = High
depression) to members of both groups. We found that those who were given the
cactus scored an average of 17.03, those who were given a bamboo plant had an
average score of 16.37, and those who were given no plant scored an average of
15.69.; using a .05 level, we found no statistically significant differences among any of
these groups (p ranged from .188 to .766). We found that these plants were not helpful
in reducing depression among these nursing home residents.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
27
(a)
H0: Eating chocolate does not enhance mood.
H1: Eating chocolate enhances mood.
(b)
Histograms with normal curve plots show a normal distribution of mood for all groups as
shown in the three figures below; hence, the pretest criterion of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
28
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
29
The homogeneity of variance score shows a significance (p) of .403; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below, is 25 for each group;
the ANOVA test becomes more robust when the ns are at least 30, but we’ll proceed
with the figures as is.
(c)
The ANOVA test revealed the following:
Descriptives
mood
ANOVA
mood
Multiple Comparisons
mood
Tukey HSD
Groups (μ = Mood) p
μ(No chocolate) = 73.00 : μ(Chocolate 1 per meal) = 76.68 .011*
μ(No chocolate) = 73.00 : μ(Chocolate 2 per meal) = 74.72 .354
μ(Chocolate 1 per meal) = 76.68 : μ(Chocolate 2 per meal) = 74.72 .262
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
30
The mean mood for Group 1 (No chocolate) is 73.00, whereas the mean mood for
Group 2 (1 per meal) is 76.68. This 3.68-point difference is statistically significant since
the significance (p) is .011 (which is less than the .05 level), so I would reject H0 and
not reject H1. No statistically significant difference (p = .354) in mood was detected
between Group 1 (No chocolate; = 73.00) and Group 3 (2 chocolates with each meal;
= 74.72).
(d)
This study analyzed the effects that chocolate with each meal has on mood. The
participants were randomly assigned to one of three groups: a group that ate their
regular meals (with no chocolate), a group that ate one piece of chocolate after
breakfast, lunch, and dinner, and a group that ate two pieces of chocolate with their
three daily meals. Results revealed a mean mood score of 73.00 on the Acme Mood
Scale for those who did not have chocolate and 76.68 for those who had one piece of
chocolate with their meals. Using a .05 level, the p value of .011 suggests that eating
one piece of chocolate with meals facilitates a significant increase in mood; hence, we
reject H0 and do not reject H1. The group that ate two pieces of chocolate with each
meal had a mean score of 74.72, however, compared to the control group, this more
modest difference is not statistically significant (p = .354). It seems that mood is
enhanced by a moderate level of chocolate in the diet administered at breakfast, lunch
and dinner among these participants.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
31
(a)
H0: Eating chocolate does not enhance mood.
H1: Eating chocolate enhances mood.
(b)
Histograms with normal curve plots show a normal distribution of mood for all groups as
shown in the three figures below; hence, the pretest criterion of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
32
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
33
The homogeneity of variance score shows a significance (p) of .060; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below, is under 30 for each
group; the ANOVA test becomes more robust when the ns are at least 30, but we’ll
proceed with the figures as is.
(c) The ANOVA test revealed the following:
Descriptives
mood
ANOVA
mood
Multiple Comparisons
mood
Sidak
Groups (μ = mood) p
μ(No chocolate) = 78.82 : μ(Chocolate 1 per meal) = 79.17 .997
μ(No chocolate) = 78.82 : μ(Chocolate 2 per meal) = 82.22 .229
μ(Chocolate 1 per meal) = 79.17 : μ(Chocolate 2 per meal) = 82.22 .361
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
34
The Sidak post hoc test was run since the groups had different ns. The Sig(nificance),
or p level, rendered on the ANOVA table is .170, which is greater than the .05 level,
indicating that there are no statistically significant differences among these (three)
groups. This is confirmed by inspecting the Multiple Comparisons table; the Sig. (p
levels) range from .229 to .997—all of which are greater than .05. While the means
varied among the groups, they are not different enough from each other to say that any
group outperformed any of the other groups. Hence, I would not reject H0 and I would
reject H1.
(d)
This study analyzed the effects that chocolate with each meal has on mood. The
participants were randomly assigned to one of three groups: a group that ate their
regular meals (with no chocolate), a group that ate one piece of chocolate after
breakfast, lunch, and dinner, and a group that ate two pieces of chocolate with their
three daily meals. Results revealed a mean mood score of 78.82 on the Acme Mood
Scale for those who did not have chocolate, 79.17 for those who ate one piece of
chocolate with their three meals per day, and 82.22 among those who ate two pieces
with each of their three meals. The mood scores of these three groups were processed
using ANOVA, which revealed a p value of .170; since this is greater than the .05
level, this indicates that despite the moderate variability in mood detected among these
groups, these differences are not considered to be statistically significant using an
level of .05. It seems that chocolate, at these levels, does not serve to enhance mood
among these participants.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
35
(a)
H0: Providing flu shot informational media has no impact on flu shot receptivity.
H1: Providing flu shot informational pamphlet has a positive impact on flu shot
receptivity.
(b)
The histograms (below) for all groups show normal distributions of flu shots, hence the
pretest criterion of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
36
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
37
The homogeneity of variance score shows a significance (p) of .428; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below, is over 30 for each
group, which satisfies that criterion.
(c)
The ANOVA test revealed the following:
Descriptives
shots
95% Confidence Interval for Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
Nothing 40 16.52 4.151 .656 15.20 17.85 7 25
Flu shot pamphlet 40 16.55 3.651 .577 15.38 17.72 9 23
Flu shot video 40 19.60 4.634 .733 18.12 21.08 10 29
Total 120 17.56 4.377 .400 16.77 18.35 7 29
ANOVA
shots
Multiple Comparisons
shots
Tukey HSD
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
38
The Tukey post hoc test was run since the groups all had the same ns (n = 40). The
Sig(nificance), or p level, rendered on the ANOVA table is .001, which is less than the
.05 level, indicating that there are statistically significant differences among these
(three) groups. This is confirmed by inspecting the Multiple Comparisons table; upon
reviewing the mean number of flu shots given per day for each group, the flu shot video
(μ = 19.60) statistically significantly outperformed no media (μ = 16.52) and the flu shot
informational pamphlet (μ = 16.55) with p = .004 ( = .05) for both comparisons. Hence,
I would reject H0 and not reject H1.
(d)
In an effort to discover if the media was effective in promoting flu shot receptivity among
patients, individuals at a walk-in health clinic were randomly assigned to one of three
groups: Group1 constituted the control group, and received no media or messaging
promoting the utility of the flu shot (which is available to all who request it at this clinic),
members of Group 2 were issued a flu shot informational pamphlet, and members of
Group 3 were shown a brief flu shot informational video, containing the same
educational points as the pamphlet. Group 1 (No information) and Group 2 (Pamphlet)
rendered an average of 16.52 and 16.55 flu shots per day respectively, whereas Group
3 (Video) statistically significantly outperformed both other groups with an average of
19.60 flu shots per day (p = .004 for both comparisons; = .05). These findings suggest
that giving a pamphlet is essentially equivalent to giving nothing (Group 1 : Group 2
rendered p = 1.00), whereas this cohort seemed most receptive to a brief informational
video promoting the flu shot. Our future studies will focus on assessing the
effectiveness of such videos by varying parameters such as duration, content, graphics,
and complexity of message.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
39
(a)
H0: Providing flu shot informational media has no impact on flu shot receptivity.
H1: Providing flu shot informational pamphlet has a positive impact on flu shot
receptivity.
(b)
The histograms (below) for the groups show normal distributions of flu shots, hence the
pretest criterion of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
40
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
41
The homogeneity of variance score shows a significance (p) of .310; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below, is over 30 for each
group, which satisfies that criterion.
(c)
The ANOVA test revealed the following:
Descriptives
shots
95% Confidence Interval for Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
Nothing 60 29.70 10.179 1.314 27.07 32.33 9 48
Flu shot pamphlet 60 33.97 8.326 1.075 31.82 36.12 16 53
Flu shot video 60 35.23 8.996 1.161 32.91 37.56 14 55
Total 180 32.97 9.450 .704 31.58 34.36 9 55
ANOVA
shots
Multiple Comparisons
shots
Tukey HSD
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
42
The Tukey post hoc test was run since the groups all had the same ns (n = 60). The
Sig(nificance), or p level, rendered on the ANOVA table is .003, which is less than the
.05 level, indicating that there are statistically significant differences among these
(three) groups. This is confirmed by inspecting the Multiple Comparisons table; upon
reviewing the mean number of flu shots given per day for each group, we find that the
pamphlet group (μ = 33.97) and the video group (μ = 35.23) statistically significantly
outperformed the group that got no media (μ = 29.70) with p = .032 and p = .003
respectively ( = .05). Hence, I would reject H0 and not reject H1.
(d)
In an effort to discover if educational media was effective in promoting flu shot
receptivity among patients, individuals at a walk-in health clinic were randomly assigned
to one of three groups: Group1 constituted the control group, and received no media or
messaging promoting the utility of the flu shot (which is available to all who request it at
this clinic), members of Group 2 were issued a flu shot informational pamphlet, and
members of Group 3 were shown a brief flu shot informational video, containing the
same educational points as the pamphlet. Group 1 (No information) (μ = 29.70 flu shots
per day) was outperformed by Group 2 (Pamphlet) (μ = 29.70, p = .032) and Group 3
(Video) (μ = 35.23, p = .003). Although the video faired best, we detected no statistically
significant difference (1.26 tests daily) between the pamphlet and the video when it
came to average daily testing. This suggests that paper or video messaging perform
essentially equivalently, and both are more effective than providing no such media. Our
future research will focus on variations on this media to assemble optimal and cost
effective messaging most suitable to our patient population.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
43
(a)
H0: A laugh track on a comedy video does not enhance viewer enjoyment.
H1: A laugh track on a comedy video enhances viewer enjoyment.
(b)
Histograms with normal curve plots show a normal distribution of enjoyment for all
groups as shown in the three figures below; hence, the pretest criterion of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
44
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
45
The homogeneity of variance score shows a significance (p) of .126; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below, is over 30 for each
group, which satisfies that criterion.
(c)
The ANOVA test revealed the following:
Descriptives
enjoy
ANOVA
enjoy
Multiple Comparisons
enjoy
Sidak
Groups (μ = enjoyment) p
μ(No laugh track) = 3.47 : μ(Laugh track 50) = 4.21 .090
μ(No laugh track) = 3.47 : μ(Laugh track 100) = 4.48 .010*
μ(Laugh track 50) = 4.21 : μ(Laugh track 100) = 4.48 .805
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
46
The Sidak post hoc test was run since the groups had different ns. The Sig(nificance),
or p level, rendered on the ANOVA table is .010, which is less than the .05 level,
indicating that there are statistically significant differences among these (three) groups.
This is confirmed by inspecting the Multiple Comparisons table; the Sig. (p levels); the
only pair that produced a statistically significant difference (p = .010) is the comparison
between Group 1 (No laugh track) and Group 3 (Laugh track with a 100-person
audience). Hence, I would reject H0 and not reject H1.
(d)
This study analyzed the effects that a laugh track has on viewer enjoyment of a comedy
video. The participants were randomly assigned to one of three groups: a group that
watched a comedy video with no laugh track, a group that watched the same video with
the laughter of a 50-member audience, and a group that watched the same video with
the laughter of a 100-member audience. Results revealed an enjoyment level of 3.47
(on a 5-point scale) for those who watched without the laugh track, 4.21 for those who
watch with the 50-person laugh track, and 4.48 for those who watch with the 100-person
laugh track. Although participants reported higher levels of enjoyment when exposed to
the 50-person laugh track, the 100-person laugh track was found most enjoyable of the
three groups. Comparing the control group to the group that watched with the 100-
person audience laugh track produced a statistically significant difference in enjoyment
(p = .010, = .05). This suggests that the laugh track significantly enhances enjoyment,
but only if there the (recorded) audience is of a sufficient size.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
47
(a)
H0: A laugh track on a comedy video does not enhance viewer enjoyment.
H1: A laugh track on a comedy video enhances viewer enjoyment.
(b)
Histograms with normal curve plots show a normal distribution of enjoyment for all
groups as shown in the three figures below; hence, the pretest criterion of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
48
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
49
The homogeneity of variance score shows a significance (p) of .694; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
Only one of the groups has an n of at least 30; the findings would be considered more
robust if the n for each group was at least 30.
(c)
The ANOVA test revealed the following:
Descriptives
enjoy
95% Confidence Interval for Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
No laugh track 35 2.49 .853 .144 2.19 2.78 1 4
Laugh track 50 26 2.62 1.098 .215 2.17 3.06 1 6
Laugh track 100 27 2.52 .975 .188 2.13 2.90 1 4
Total 88 2.53 .958 .102 2.33 2.74 1 6
ANOVA
enjoy
Multiple Comparisons
enjoy
Sidak
Groups (μ = Enjoyment) p
μ(No laugh track) = 2.49 : μ(Laugh track 50) = 2.62 .939
μ(No laugh track) = 2.49 : μ(Laugh track 100) = 2.52 .999
μ(Laugh track 50) = 2.62 : μ(Laugh track 100) = 2.52 .977
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
50
The Sidak post hoc test was run since the groups had different ns. The Sig(nificance),
or p level, rendered on the ANOVA table is .871, which is greater than the .05 level,
indicating that there are no statistically significant differences among these (three)
groups. This is confirmed by inspecting the Multiple Comparisons table; the Sig. (p
levels), which are all greater than .05. Based on these findings, I would not reject H0
and reject H1.
(d)
This study analyzed the effects that a laugh track has on viewer enjoyment of a comedy
video. The participants were randomly assigned to one of three groups: a group that
watched a comedy video with no laugh track, a group that watched the same video with
the laughter of a 50-member audience, and a group that watched the same video with
the laughter of a 100-member audience. Results revealed an enjoyment level of 2.49
(on a 5-point scale) for those who watched without the laugh track, 2.62 for those who
watch with the 50-person laugh track, and 2.52 for those who watch with the 100-person
laugh track. Comparing these three groups to each other revealed no statistically
significant differences (p > .93 for all three groups; = .05). These findings suggest that
the laugh track does not significantly enhance enjoyment among these participants,
hence, we do not reject our null hypothesis (H0), and reject H1, which states that
including a laugh track will enhance audience enjoyment.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
51
(a)
H0: Light therapy has no effect on depression.
H1: Light therapy is effective in reducing depression.
(b)
Histograms with normal curve plots show a normal distribution of mood for all groups as
shown in the three figures below, hence, the pretest criteria of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
52
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
53
The homogeneity of variance score for mood shows a significance (p) of .246; since this
is greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the three groups, hence, this pretest criterion
passes.
The n for each group is 80 in each group, which satisfies the 30 per group minimum
criterion (see Descriptives table below).
(c)
The ANOVA test revealed the following:
Descriptives
mood
ANOVA
mood
Multiple Comparisons
mood
Tukey HSD
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
54
Groups (μ = mood) p
μ(No light th.) = 59.1 : μ(Light th. even days) = 61.9 .053
μ(No light th.) = 59.1 : μ(Light th. every day) = 62.7 .008*
μ(Light th. even days) = 61.9 : μ(Light th. every day) = 62.7 .770
Means rounded to one decimal digit.
*Statistically significant difference ( = .05).
After 1 month of treatment, participants who received light therapy for 1 hour a day
scored an average of 62.7 on a mood test, compared to 59.1 among those who had no
light therapy; this 3.6 difference in their scores is statistically significant (p = .008, =
.05). Those who received 1 hour of light therapy every other day showed a better mood
score (61.9) than those who received no light therapy, however, the improvement is not
considered to be statistically significant (p = .053) using the .05 level. Incidentally,
comparing the mean mood score of those who received light therapy (61.9) for every-
other-day, and for daily use (62.7) revealed no statistically significant difference
between those two groups (p = .770, = .05).
(d)
In order to determine if light therapy is a viable supplement to treating depression, 240
subjects diagnosed with depression were randomly assigned to one of three groups:
The control group received no light therapy; the second group received 1 hour of light
therapy every other day for a month; the third group received light therapy for 1 hour
every day for a month. After 30 days, all participants completed the Acme Mood Scale,
a 10 question survey that renders a score from 1 to 100 (1 = Extremely bad mood, 100
= Extremely good mood). The group average scores on the mood test were similar
among the two light therapy groups—61.9 for the every-other day group, and 62.7 for
the daily group. Although these differences were not statistically significantly different
from each other (p = .770, = .05), only the group that received light therapy on a daily
basis (μ = 62.7) showed a statistically significant improvement in mood, when compared
to those who received no light therapy (μ = 59.1) (p = .008, = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
55
(a)
H0: Light therapy has no effect on depression.
H1: Light therapy has is effective in reducing depression.
(b)
Histograms with normal curve plots show a normal distribution of mood for all groups
per the three figures below, hence, the pretest criterion of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
56
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
57
The homogeneity of variance score for mood shows a significance (p) of .262; since this
is greater than the level of .05, this suggests that there is no statistically significant
difference between the variances among the three groups, hence, this pretest criterion
passes.
The ns for these groups are 48, 56, and 37, which satisfies the 30 per group minimum
criterion (see Descriptives table below).
(c)
The ANOVA test revealed the following:
Descriptives
mood
ANOVA
mood
Multiple Comparisons
mood
Sidak
Groups (μ = mood) p
μ(No light th.) = 45.44 : μ(Light th. even days) = 44.86 .843
μ(No light th.) = 45.44 : μ(Light th. every day) = 45.54 .999
μ(Light th. even days) = 44.86 : μ(Light th. every day) = 45.54 .805
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
58
After 1 month of treatment, participants who received light therapy for 1 hour a
day scored an average of 45.54 on a mood test; those who had light therapy every
other day scored an average of 44.86, and those who had no light therapy had an
average score of 45.44. The ANOVA table reports a Sig. (p) value of .657 which is
greater than the .05 level, indicating that there are no statistically significant
differences among the groups. This is confirmed by the Sig. column on the Multiple
Comparisons table, wherein the p values range from .805 to .999, which is well above
the .05 level. Hence, we can conclude that for this sample, the light therapy protocol
produced no statistically significant improvement in mood.
(d)
In order to determine if light therapy is a viable supplement to treating depression, 141
subjects diagnosed with depression were randomly assigned to one of three groups:
The control group received no light therapy; the second group received 1 hour of light
therapy every other day for a month; the third group received light therapy for 1 hour
every day for a month. After 30 days, all participants completed the Acme Mood Scale,
a 10 question survey that renders a score from 1 to 100 (1 = Extremely bad mood, 100
= Extremely good mood). The group average scores on the mood test were similar;
45.44 for those who received no light therapy, 44.86 for those who had light therapy
every other day, and 45.54 for those who had light therapy daily. The p values among
these three groups ranged from .805 to .999; using an level of .05, we conclude that
this schedule of light therapy did not produce a statistically significant improvement in
mood for these participants.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
59
(a)
H0: Walking in the morning will have no effect on energy.
H1: Walking in the morning enhances energy.
(b)
Although the histogram for Group 1 (below) shows some low outliers, the majority of the
data forms a normal curve for energy. The histograms for the other groups appears
normally distributed. I’d say the pretest criterion of normality is sufficiently satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
60
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
61
The homogeneity of variance score shows a significance (p) of .134; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below are greater than 30; the
n quotas are satisfied.
(c)
The ANOVA test revealed the following:
Descriptives
energy
95% Confidence Interval for Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
No walking 132 81.23 4.328 .377 80.49 81.98 61 88
Walking: 30 min. 124 82.60 4.320 .388 81.84 83.37 73 93
Walking: 60 min. 95 82.77 4.716 .484 81.81 83.73 73 93
Total 351 82.13 4.476 .239 81.66 82.60 61 93
ANOVA
energy
Multiple Comparisons
energy
Sidak
The Sidak post hoc test was used since the ns for each group differed from each other.
Groups (μ = energy) p
μ(No walking) = 81.23 : μ(Walking 30 minutes) = 82.60 .041*
μ(No walking) = 81.23 : μ(Walking 60 minutes) = 82.77 .031*
μ(Walking 30 minutes) = 82.60 : μ(Walking 60 minutes) = 82.77 .990
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
62
End-of-the-day energy was statistically significantly higher for the groups that walked
compared to the group that did not walk. These findings lead us to reject H0 and not
reject H1.
(d)
This study measured the end-of-day energy of individuals assigned to one of three
groups; Group 1 served as the control group and did no walking; the other groups took
a morning walk—Group 2 walked for 30 minutes, and Group 3 walked for 1 hour. At the
conclusion of this study, participants answered the Acme End-of-the-Day Energy Scale.
This instrument produces a score between 1 and 100 (1 = Extremely low energy, 100 =
Extremely high energy). Our analysis showed that those who did not walk scored 81.23;
those who walked had statistically significantly higher levels of energy using an level
of .05—Group 2 scored 82.60 (p = .041), and those in Group 3 scored 82.77 (p = .031).
Although Group 3 scored slightly higher than Group 2, this is not considered to be a
statistically significant difference (p = .990). Based on these findings, we reject H0 which
asserts that that walking made no difference in one’s energy, and we do not reject H1.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
63
(a)
H0: Walking in the morning will have no effect on energy.
H1: Walking in the morning enhances energy.
(b)
The histogram for the three groups (below) appear to be normally distributed for energy;
the pretest criterion of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
64
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
65
The homogeneity of variance score shows a significance (p) of .456; since this is
greater than the level of .05, this suggests that there is no statistically significant
difference between the variances of the groups; hence, this pretest criterion passes.
The n for each group, as shown in the Descriptives table below are greater than 30; the
n quotas are satisfied.
(c)
The ANOVA test revealed the following:
Descriptives
energy
95% Confidence Interval for Mean
N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum
No walking 90 54.22 10.994 1.159 51.92 56.52 32 77
Walking: 30 min. 90 55.43 9.971 1.051 53.34 57.52 35 77
Walking: 60 min. 90 58.33 10.349 1.091 56.17 60.50 35 79
Total 270 56.00 10.550 .642 54.73 57.26 32 79
ANOVA
energy
Multiple Comparisons
energy
Tukey HSD
The Tukey post hoc test was used since the ns for each group were 90.
Groups (μ = Energy) p
μ(No walking) = 54.22 : μ(Walking 30 minutes) = 55.43 .717
μ(No walking) = 54.22 : μ(Walking 60 minutes) = 58.33 .024*
μ(Walking 30 minutes) = 55.43 : μ(Walking 60 minutes) = 58.33 .152
*Statistically significant difference ( = .05).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
66
Those in the walking groups (Groups 2 and 3) had higher levels of energy than those
who did not walk (Group 1); although those who walked for 60 minutes had higher
energy than those who walked for only 30 minutes, the energy difference between these
two groups is not statistically significant. The only significant difference was detected
between Group 1 and Group 3. These findings lead us to reject H0 and not reject H1.
(d)
This study measured the end-of-day energy of individuals assigned to one of three
groups; Group 1 served as the control group and did no walking; the other groups took
a morning walk—Group 2 walked for 30 minutes, and Group 3 walked for 1 hour. At the
conclusion of this study, participants answered the Acme End-of-the-Day Energy Scale.
This instrument produces a score between 1 and 100 (1 = Extremely low energy, 100 =
Extremely high energy). Our analysis showed that those who walked had more energy
than those who did not walk—in fact, the more walking, the more energy participants
reported, however, only those in Group 3, who walked for an hour each morning ( =
58.33) had energy scores that were statistically significantly higher than those who did
not walk ( = 54.22); this comparison produced a p value of .024 ( = .05). Based on
these findings, we reject H0, which stated that walking made no difference in one’s
energy and we would fail to reject H1.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
67
(a)
H0: Lighting source has no effect on reading rate.
H1: Lighting source has an effect on reading rate.
(b) Histograms with normal curve plots show a normal distribution of seconds for all
groups as shown in the four figures below, hence, the pretest criterion of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
68
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
69
The homogeneity of variance score for seconds shows a significance (p) of .219; since
this is greater than the level of .05, this suggests that there is no statistically
significant difference between the variances among the four groups, hence, this pretest
criterion passes.
The Generic lamp group was the only group that had the minimal n of 30. The Room
lighting, Acme lighting, and Flashlight groups had ns of 25, 28 and 24 respectively (see
Descriptives table below). The findings of the ANOVA test would be more robust if the
ns were slightly higher for these three groups.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
70
ANOVA
seconds
Multiple Comparisons
seconds
Sidak
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
71
Per the table above, since the mean reading time in the Acme reading lamp group is
statistically significantly lower than scores of those who read using Room lighting and by
Flashlight, we reject H0. For the same reason, we would not reject H1.
(d)
This study analyzed the effects that the Acme reading lamp had on reading speed. The
107 participants were randomly assigned to one of four groups; one group read a 1,000-
word essay using regular room lighting, the second group read the same essay using
the new Acme reading lamp, the third group read using a generic reading lamp, and the
fourth group read using a flashlight. Results revealed that on the average, those who
read using the Acme reading lamp read significantly faster (μ = 406) than those who
read using a flashlight (μ = 441, p = .008), or regular room lighting (μ = 436, p = .032)
using an level of .05. Incidentally, those who used a generic reading lamp (μ = 410)
finished reading the essay significantly faster than those who read by flashlight (μ =
441, p = .020). We also discovered that those who read using a Acme reading lamp (μ =
406) completed the essay faster than those who used the generic reading lamp (μ =
410), however this difference was not found to be statistically significant (p = .999).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
72
(a)
H0: Lighting source has no effect on reading rate.
H1: Lighting source has an effect on reading rate.
(b)
Histograms with normal curve plots show a normal distribution of seconds for all groups
as shown in the four figures below, hence, the pretest criterion of normality is satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
73
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
74
The homogeneity of variance score for seconds shows a significance (p) of .328; since
this is greater than the level of .05, this suggests that there is no statistically
significant difference between the variances among the four groups, hence, this pretest
criterion passes.
The groups each had an n of 25 (see Descriptives table below). The findings of the
ANOVA test would be more robust if the ns were at least 30 per group.
(c)
The ANOVA test revealed the following:
Descriptives
seconds
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
75
ANOVA
seconds
Multiple Comparisons
seconds
Tukey HSD
Per the table above, since the mean reading time in the Acme lamp group (μ = 416) is
statistically significantly lower than those who read using a Flashlight (μ = 449, p =
.022), based on the .05 level, we reject H0 and not reject H1. Additionally, Room
lighting (μ = 416) statistically significantly outperformed the reading rate of the Flashlight
(μ = 449, p = .027), and finally, the reading rate for the generic lamp (μ = 416)
outperformed the Flashlight (p = .013).
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
76
(d)
This study analyzed the effects that the Acme reading lamp had on reading speed. The
100 participants were randomly assigned to one of four groups; one group read a 1,000-
word essay using regular room lighting, the second group read the same essay using
the new Acme reading lamp, the third group read using a generic reading lamp, and the
fourth group read using a flashlight. There was no statistically significant difference in
reading times among those who used the Acme reading lamp (μ = 416), Room lighting
(μ = 416) or the Generic lamp (μ = 414) using an level of .05. All three of those groups
read statistically significantly faster than the fourth group, who read using a flashlight (μ
= 449); p values ranged from .013 to .027.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
77
(a)
H0: Classical music has no effect on problem-solving skills.
H1: Classical music enhances problem-solving skills.
(b)
Histograms with normal curve plots show a normal distribution of seconds for all of the
groups as shown in the four figures below, hence, the pretest criterion of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
78
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
79
The homogeneity of variance score for seconds shows a significance (p) of .319; since
this is greater than the level of .05, this suggests that there is no statistically
significant difference between the variances among the four groups, hence, this pretest
criterion passes.
The n for each group is 40 (see Descriptives table below), which satisfies the 30 per
group minimum criterion.
(c)
The ANOVA test revealed the following:
Descriptives
seconds
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
80
ANOVA
seconds
Multiple Comparisons
seconds
Tukey HSD
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
81
(d)
This experiment was designed to determine if classical music facilitates problem
solving. We recruited 160 people, and randomly assigned 40 participants to one of four
groups: Members of each group were given a 100-piece jigsaw puzzle to solve; those in
Group 1 worked in silence, those in Groups 2, 3, and 4 solved the puzzle with classical
music playing at different volumes in each group—soft (30 dB), medium (60 dB), and
loud (90 dB) respectively. We found that those who listened to the soft classical music
(at 30 dB) completed the puzzle significantly faster (μ = 594) than those who worked in
silence (μ = 620) (p = .047; = .05), however, it seems that as the music was played
louder, the beneficial effect was lost: Those who listened at a medium volume
completed the puzzle in 596 seconds, and those who were exposed to the loud volume
took 626 seconds. It seems that while the soft classical music helped facilitate the
problem solving, louder volumes may have been distracting.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
82
(a)
H0: Classical music has no effect on problem-solving skills.
H1: Classical music enhances problem-solving skills.
(b)
Histograms with normal curve plots show a normal distribution of seconds for all of the
groups as shown in the four figures below, hence, the pretest criterion of normality is
satisfied.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
83
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
84
The homogeneity of variance score for seconds shows a significance (p) of .284; since
this is greater than the level of .05, this suggests that there is no statistically
significant difference between the variances among the four groups, hence, this pretest
criterion passes.
The n for each group is over 30 (see Descriptives table below), which satisfies the 30
per group minimum criterion.
(c)
The ANOVA test revealed the following:
Descriptives
seconds
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
85
ANOVA
seconds
Multiple Comparisons
seconds
Sidak
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.
86
There was no statistically significant difference in the participants’ time to complete the
puzzle among those who worked in silence (μ = 609), those who listened to classical
music at a soft volume (μ = 606), and those who listened at a medium volume (μ = 608)
using an of .05; however, those who solved the puzzle with the classical music
playing loud (90 dB) took significantly longer (μ = 671) compared to those who worked
with no music (p = .015), music at a soft volume (p = .001), or moderate volume (p =
.013). It seems that classical music does not aid in such problem solving, in fact, if
played too loud, it can serve as a distraction.
(d)
This experiment was designed to determine if classical music facilitates problem
solving. We recruited 153 subjects, and randomly assigned them to one of four groups:
those in Group 1 worked on a 100-piece jigsaw puzzle in silence; those in Groups 2, 3,
and 4 solved the same puzzle with classical music playing at different volumes in each
group—soft (30 dB), medium (60 dB), and loud (90 dB), respectively. We found no
statistically significant difference in puzzle solving time among those who worked in
silence (μ = 609), or with classical music played at a soft volume (μ = 606) or moderate
volume (μ = 609) (p > .05). However, participants who worked on the puzzle while
listening to loud classical music (90 dB) took significantly longer (μ = 617) to solve the
puzzle compared to those who worked in silence (μ = 509) (p = .015, = .05),
suggesting that classical music has no effect on this sort of problem solving, unless it is
played loud, in which case, the music appears to act as a distracter.
Knapp, Introductory Statistics Using SPSS, Second Edition. © 2017, SAGE Publications.