Empiri cal Statistics Paper

You’ve done all of the statistical analyses necessary to complete your empirical paper. Now, it is time for you to get a feel for what your empirical paper should look like.

1. Review Appendix A. It contains the Statistical Analysis Process that is embedded in every empirical paper. You should be very familiar with this process by now.
2. Review Appendices B and C. Appendix B lists the general specifications for your paper. Appendix C displays the rubric that I’ll use to grade your paper. I’m showing you B and C so that you you’ll have a very good idea about what specifically I’ll be looking for in your paper.
3. Use the Zelbst et al. (2010) as a pattern. Your paper will be very similar to this one although yours will be an abbreviated form. This paper has 37 pages and 9,671 words. Your paper will only have 15 pages. I’ll want you to follow the citation and reference formatting conventions illustrated in this paper. Note that the method for assessing the measurement scales for validity is somewhat different than the one you learned. You’ll stick with what you know. Look at all of the tables and figures. Your tables and figures will look like those in this paper. The paper is posted in Blackboard.

Statistical Analysis Process:

Generally, all of the business courses that you take are focused on answering a single question - "What can managers do to improve the performance of their organizations?" It is not sufficient to theorize that a particular business strategy will lead to improved performance. You must test the theory. The statistical analysis process is a standardized methodology for testing theories. Generally, you identify a strategy that you believe may work, you ask experts what they think about the strategy, you analyze the data collected from the experts, you interpret the results from the analysis, and you draw conclusions about how well the strategy works based on the interpretations.

All remaining assignments require that you follow this process.

1. State the research question.
   This is general question related to our area of concern. For example: "What can managers do to improve the performance of their organizations?" That’s probably stated a little too generally. More specific examples include: "Will implementation of a market orientation strategy improve my organization’s performance?" and "Will implementation of JIT and TQM improvement programs combine to improve my organization’s performance?"
2. State the hypotheses.
   Hypotheses are formulated from the research question. They must be stated in a form that is statistically testable. For example, "Market orientation is positively associated with operational performance." Another example is: "JIT and TQM are positively associated with organizational performance." These statements are statistically testable using correlation and regression analyses. Correlation analysis gives the type, strength, and significance level of the association, and regression analysis provides information necessary to predict OP based on given levels of market orientation, JIT, and TQM.
3. Identify measurement scales.
   This requires that measurement scales like the 10-item scale for market orientation and the 8-item scale for operational performance be found or developed. Finding an existing scale is easy. Creating a new scale is not so easy. In this class, you’ll only be dealing with existing scales that I’ve used in my research.
4. Collect data.

Then, data is collected from a sample of respondents that have knowledge about the variables you are concerned about, like MO, JIT, TQM, and OP. In this class, with one exception, you’ll be dealing with existing data sets. There are many ways to collect data. We will focus on the collection of data through the standard survey methodology. A survey instrument is crafted with demographic questions and measurement scales, and experts (usually practicing managers) are asked to complete the surveys either by mail or e-mail. The dataset is built using the data provided by the experts. Now, you have an SPSS or Excel data file to use to complete the statistical analyses necessary to eventually answer the research question.

1. Assess the measurement scales for validity and reliability.
   Validity means that the scales measure what they are supposed to measure. We are concerned about three types of validity: content (face), discriminant, and convergent. Content validity cannot be statistically measured. You just have to look at the items in the scale and make certain that each item is focused on measuring what you want to measure. Discriminant validity means that each measurement scale is measuring something different. For example, the items in the market orientation scale are measuring something (market orientation) distinctly different than any other scale (say operational performance). Convergent validity means that all of the items in a scale are measuring the same thing. For example, the items in the market orientation scale don’t overlap with any of the items in the operational performance scale. Discriminant and convergent validity can be measured using confirmatory factor analysis. If the market orientation items load on a single factor and the items in the operational performance scale all load on a separate single factor, the scales exhibit both discriminant and convergent validity.
   Reliability means that the measurement scales measure consistently across similar samples. It is measured with Cronbach’s alpha. If the alpha value for a scale is greater than or equal to .70, the scale is considered to be sufficiently reliable.
   So, the statistical analyses necessary at this step include confirmatory factor analysis and reliability analysis, both of which SPSS does very nicely. If you conclude that the measurement scales are sufficiently valid and reliable, you can proceed. If the scales aren’t valid and reliable, you’ve got to stop because subsequent analyses will be meaningless.
2. Compute summary variables.
   This is a relatively simple step. Although not the only way to accomplish the task, most often the values that respondents record for the individual items in a scale are averaged to get a mean value called the summary variable. For example, values for the ten market orientation items (MO1 through MO10) are averaged to get MO. SPSS does this easily through its [transform, compute variable] capability.
3. Conduct statistical analyses.
   This is where the statistics (descriptive, correlation, and regression) are actually generated. The descriptive statistics table tells us if the summary variables are normally distributed. Normality is a necessary condition for us to conduct correlation and regression analyses. The correlation matrix tells us the type, strength and significance level of the correlation coefficients (R). Regression analysis tells us the coefficient of determination (R ² ) and the regression coefficients for use in predicting the dependent variable given specific values of the independent variables. The R ² and beta values from the regression results are used to complete path analysis
4. Interpret results.
   This step requires that you interpret the results of the statistical analyses. What do the statistics generated actually mean? How do you know whether the measurement scales are valid and reliable? How do you know that the summary variables are normally distributed? How do you know what kind of relationship there is between the independent and dependent variables?
5. Draw conclusions and state recommendation for practitioners.

Finally, and this is generally the most difficult step, draw a conclusion based on the results and your interpretation of the results. This means answer the research question and summarize the evidence supporting your answer. For example, "Implementation of a market orientation strategy will improve the performance of your organization." This conclusion leads to a recommendation that "managers implement a market orientation strategy." Your conclusion and recommendation are based on the finding of a positive, significant relationship between market orientation and performance.

Price: $35.31

Solution: The downloadable solution consists of 18 pages, 1731 words and 2 charts.
Deliverable: Word Document