The Effects of Statistical Analysis Software and Calculators on Statistics Achievement Paris Interna

1
The Effects of Statistical Analysis Software
andCalculators on Statistics AchievementParis
International Conference on Education, Economy
SocietyNovotel Paris Tour Eiffel, 17 July 2008

• Edwin P. Christmann Slippery Rock University
  USA
• edwin.christmann_at_sru.edu

2
Prior ResearchChristmann, E.P., Badgett, J.L.
(1997b). The Effects of Computers and Calculators
on Male and Female Statistics Achievement.
Journal of Computing in Higher Education, 9(1),
49-58.

• Christmann and Badgett (1997b) did not find any
  difference in academic achievement between
  students who used calculators and
  microcomputer-based statistical analysis
  software, they did find an interaction effect
  showing that females performed better on
  statistical calculations when using handheld
  calculators as opposed to using
  microcomputer-based statistical analysis
  software. Therefore, this study seeks to once
  again explore the effects of statistical analysis
  software and calculators on the statistics
  achievement of male and female university
  students.

3
This study has thus focused on the following two
null hypotheses

• 1. There will be no significant difference
  between the achievement of university students
  who used microcomputer-based software and those
  who used handheld calculators in a univariate
  statistics course
• 2. There will be no significant difference
  between the achievement of male and female
  university students who used microcomputer-based
  software and those who used handheld calculators
  within a univariate statistics course.

4
Methodology

• Subjects
• The participants were 73 graduate students
  enrolled in a univariate statistics course at a
  northeastern public, comprehensive university
  enrolling approximately 8100 students. The
  subjects were 18 males and 55 females who were
  enrolled in a required, one-semester graduate
  statistics course.

5
Methodology (continued)

• Intervention
• Microcomputers. Macintosh and PC computers were
  used by the students in the microcomputer group.
• Computer Software. GraphPad Statistical
  Applications software was utilized by the
  students in the microcomputer group. This
  software was used to calculate means, minimums,
  maximums, sums, standard deviations, variances,
  standard errors of the mean, and ranges.
• Handheld Calculator. The Texas Instruments TI-30X
  Scientific Calculator was used by the students in
  the handheld calculator group. The calculators
  were used to calculate means, minimums, maximums,
  sums, standard deviations, variances, standard
  errors of the mean, and ranges.
• Dependent Measures. The dependent measure was a
  teacher-designed test that was edited by the
  research staff on the basis of course unit
  objectives. The test contained 50 questions that
  assessed the students' abilities to analyze and
  interpret statistical data. The Spearman-Brown
  Prophecy Formula was used in calculating the
  reliability of the test. The reliability
  coefficient was calculated to be 0.912.

6
Methodology (continued)

• Research Design
• The study employed Gall, Borg, and Gall's (1996)
  description of the Randomized Posttest Only
  Control Group Design.

7
Results Statistics Software versus Calculators

• Two-way ANOVA results yielded a significant
  difference between the computing methods between
  those students who used microcomputer-based
  statistics software and handheld calculators
  F(l,69) 5.317, p .024, with the group
  exposed to microcomputer-based statistics
  software scoring significantly higher than those
  exposed to handheld calculators. The overall mean
  posttest score for the group using
  microcomputer-based statistics software was 86.92
  (SD 8.57), while the overall mean posttest
  score for the group using handheld calculators
  was 83.93 (SD 8.35).

8
Results Gender

• The two-way ANOVA yielded no significant posttest
  results with respect to gender F(l,69) 0.245,
  p .622. However, a significant interaction
  effect was disclosed (p 0.027), with an effect
  size of 0.621 indicating that, on the average,
  those males who used microcomputers outperformed
  73 of the females who used microcomputers in
  performing their statistical calculations while
  the females who used calculators outperformed 71
  of the male users of calculators, on the basis of
  an effect size calculation of 0.545.

9
Conclusions

• The results of this research disclose a
  significant difference between the mean
  achievement scores of students who used
  microcomputer-based statistical analysis software
  and those who used handheld calculators in a
  univariate statistics course.

10
Conclusions

• When it comes to computer applications, it is
  evident that in 2008, a higher proportion of
  students appear more knowledgeable than they did
  in 1997.
• Subsequently, it is possible that students find
  it easier to use microcomputer-based statistical
  analysis software than in the past.

11
Conclusions

• Although the results reveal no significant
  differences between the scores of males and
  females who used microcomputers or handheld
  calculators in the solution of statistical
  problems, a significant interaction effect
  reveals that 73 of those male students who used
  microcomputer-based statistics software scored
  higher than the females using this software. Yet,
  as disclosed by an effect size calculation, those
  females who used handheld calculators
  outperformed 71 of the male users of the
  calculators.

12
Conclusions

• These figures definitely reflect the necessity
  for continuing research regarding the
  effectiveness of microcomputer-based statistical
  analysis software and handheld calculators on
  university statistics achievement. Undoubtedly,
  additional research is imperative for determining
  the most effective and efficient means of
  bolstering the achievement of students enrolled
  in statistics courses.
• For example, the cost of a calculator is about
  12.00 however, the cost of a microcomputer-based
  statistical software station is about 2,000.00.
  Therefore, the question must be asked
• "What is the most cost-effective technological
  tool for statistics instruction?"