Data Interpretation Practicum
Run either t tests or ANOVA on your chosen data. This Application requires you to engage in data
interpretation and to select the appropriate analyses for your hypotheses and for the data that you have at
your disposal. Toward that end, you should consider which analyses will inform the reader and allow you
to pursue your questions
your narrative interpretation; the governing assumptions of the analyses you ran; the viable and nonviable
hypotheses (null and alternative); and the relevant values (such as a P value indicating statistical
significance or a lack thereof).
Be sure to indicate to your Instructor why you selected the analyses that you did. In other words, why did
you select t test over ANOVA or vice versa? Why one-way or two-way? How is this analysis related to the
hypothesis?
DATA INTERPRETATION PRACTICUM 2
One-way analysis of variance (ANOVA)
The one-way analysis of variance (ANOVA) is a powerful data analysis tool which is
used in determining if there are any significant differences between the means of two or more
independent groups (even though what is often used in cases where the minimum number of
groups is three, instead of two groups). For instance, in this data interpretation practicum the
one-way ANOVA will used to carry out the data analysis with an aim of understanding whether
injury rate differed based on the site of work, dividing workers into three independent groups
(e.g., Boston, Phoenix and Seattle). In addition, it is imperative to recognize that the one-way
ANOVA is an omnibus test statistic meaning that it cannot provide information on the specific
independent groups with significant differences from each other, but only provides information
on at least two significantly different groups.
The governing assumptions of the one-way ANOVA include: (1) the dependent variable
measurement should be done at the ratio or interval level meaning they are continuous; (2) there
should be two or more categorical independent groups of the independent variable; (3) the
observations should be independent; (4) there should be no significant outliers; (5) there should
be normal distribution of the dependent variable for each independent variable category; and (6)
there should be homogeneity of variances. The data analysis and interpretation is guided by the
proposed null and alternative hypothesis. For example, considering that that study variable in this
task is the injury rates across three independent work sites (e.g., Boston, Phoenix and Seattle),
then one-way ANOVA can be conducted to determine if there significant differences in injury
rates across the three work sites. Thus, the null and alternative hypotheses are stated as follows:
Null hypothesis (H 0 ): mean boston = mean phoenix = mean seattle
Alternative hypothesis (H A ): mean boston ≠ mean phoenix ≠ mean seattle
DATA INTERPRETATION PRACTICUM 3
SPSS Output
NMeanStd. DeviationStd. ErrorLower BoundUpper Bound
Boston1515.629313.87773.58321.751629.507
Phoenix1917.177421.20514.8648-4.027738.3825
Seattle1712.537616.35583.9669-3.818228.8934
Total5115.175517.47452.4469-2.03143332.26097
95% Confidence Interval for Mean
Descriptives
Injury Rate
ANOVA Summary
Injury Rate
Sum of squaresdfMean squareFSig.
Between groups197.5221298.7610.310.734904
Within groups15070.334248313.9653
Total15267.856350
The null hypothesis stated as H 0 : mean boston = mean phoenix = mean seattle ; is not rejected because the
overall F statistic was not significant [F(2, 15) = 0.31, p > .05]. This indicates that the injury rate
means across the three work sites are not significantly different. The reason why ANOVA was
chosen over t-test is that it enables comparison of more than two means making it the appropriate
choice because the dataset consisted of three independent group means (i.e. Boston, Phoenix and
Seattle).
DATA INTERPRETATION PRACTICUM 4
References
Bernard, H. R. (2000). Social research methods: Qualitative and quantitative approaches.
Thousand Oaks, CA: Sage Publications.
Boslaugh, S., & Watters, P. A. (2008). Statistics in a Nutshell – Research Design. Sebastopol,
California: O’Reilly Media, Inc.
Campbell, D. T., & Stanley, J. (2010). Experimental and quasi-experimental designs for
research, (Laureate Education, Inc., custom ed.). Mason, OH: Cengage Learning.
Creswell, J. W. (2008). Research design: Qualitative, quantitative, and mixed methods
approaches, (3 rd ed.). Thousand Oaks, CA: Sage.
Green, S. B., & Salkind, N. J. (2008). Using SPSS for Windows and Macintosh: Analyzing and
Understanding Data, (6 th ed.). Upper Saddle River, NJ: Pearson Publishers.
