Case Chapter 15 Quantitative Data Analysis The Subway Family

Case Chapter 15 Quantitative Data Analysisthe Subway Familyin 1965 A

Analyze the process of quantitative data analysis based on a case involving Subway franchisees. The case provides a questionnaire measuring variables such as cohesion, recognition, autonomy, pressure, fairness, and satisfaction with the franchisor. The task involves discussing preliminary data analysis steps, measures of central tendency and dispersion, issues with data measurement, visualization techniques, reliability testing with Cronbach’s alpha, and examining relationships between variables.

Specifically, explain the initial steps before data analysis, recommend appropriate statistical measures for understanding individual items and sample characteristics, identify potential problems with how age and franchise duration are measured, describe histograms and box-and-whisker plots, evaluate the suitability of visualizations for all items, clarify reverse scoring and its application in reliability testing, interpret Cronbach’s alpha results for the cohesion measure, and suggest methods to analyze relationships among the constructs.

Sample Paper For Above instruction

Introduction

Quantitative data analysis is an essential component of research in business management, enabling researchers to interpret numerical data systematically and draw meaningful conclusions. In the context of Subway franchisees in Europe, analyzing questionnaire data measuring constructs such as cohesion, recognition, autonomy, pressure, fairness, and satisfaction with the franchisor provides insights into factors influencing franchisee satisfaction and compliance. This paper discusses the preliminary steps in data analysis, explores appropriate descriptive measures, addresses potential measurement issues, examines visualization techniques, evaluates reliability testing procedures, and outlines methods for analyzing relationships between variables.

Preliminary Steps in Data Analysis

Before performing formal analyses, several preparatory steps are essential. These include data cleaning (checking for missing or inconsistent responses), coding the data accurately (ensuring that Likert-scale responses are numerically represented), and verifying data entry accuracy. Additionally, it is crucial to screen the data for outliers and to assess the distribution of responses to determine the suitability of subsequent parametric analyses. Data validation ensures that the dataset accurately reflects respondents’ answers and that any anomalies do not distort results (Pallant, 2016).

Measures of Central Tendency and Dispersion

Items Measuring Dependent and Independent Variables

For each item in the questionnaire measuring constructs like cohesion, recognition, autonomy, pressure, and fairness, the most suitable measures of central tendency are the mean and median. These statistics summarize the typical response and provide an overview of respondents’ sentiments. The mean is especially appropriate given the ordinal Likert scale responses, assuming the data approximates interval properties. To assess variability within responses, standard deviation and range are appropriate dispersion measures, revealing how responses spread around the central point (Field, 2013).

Sample Characteristics

For demographic variables such as age and duration as a franchisee, the mode can be useful for categorical data, while the mean and standard deviation for numerical summaries. Since age and franchise duration are continuous variables, mean and standard deviation provide insight into the average respondent's profile and the variation within the sample. These measures help contextualize the other analyses by describing the sample population (Allen & Seaman, 2007).

Potential Problems with Data Measurement

The measurement of age and franchise duration using broad categorical ranges may lead to reduced data precision, which can affect subsequent analyses. These categories do not capture exact values, leading to potential loss of variability and bias in estimating relationships. Furthermore, respondents might interpret categories differently, inducing measurement error. Precise measurement methods, such as recording actual ages and exact years as a franchisee, would improve data quality and analytic accuracy (Howell, 2013).

Visualization Techniques: Histograms and Box-and-Whisker Plots

A histogram visually displays the frequency distribution of a continuous variable, revealing the shape (e.g., skewness, modality) of the data (Everitt & Hothorn, 2011). It is effective for understanding the distribution of responses and detecting outliers or data irregularities.

A box-and-whisker plot summarizes the distribution, showing the median, quartiles, and potential outliers. It provides a quick visual comparison of different variables or groups and helps assess data symmetry.

Applying these plots to all items may be insightful but could be resource-intensive. It is prudent to prioritize variables with skewed distributions or outliers for detailed visualization while using summary statistics for others. For a large set of items, focusing on key variables ensures efficient analysis and clearer interpretation (Tukey, 1977).

Reverse Scoring and Reliability Measures

Reverse scoring involves transforming responses to certain items so that higher scores consistently reflect higher levels of the underlying construct. It corrects for reversed worded items, ensuring internal consistency measures accurately reflect the construct (Neuman, 2014).

In the questionnaire, items assessing perceived fairness and recognition might require reverse scoring if they are negatively worded. For example, if an item states “I do not feel recognized,” an agreement indicates low recognition, which must be reversed to align with positive recognition scores (DeVellis, 2016).

Inter-Item Reliability Analysis

The Cronbach’s alpha for the cohesion measure is 0.804, indicating good internal consistency (Nunnally & Bernstein, 1994). The analysis of individual items suggests that most items correlate well with the total scale, with the exception of item cohesion2, which has a lower corrected item-total correlation, indicating potential issues with that item.

To establish a reliable measure, the scale should have a Cronbach’s alpha above 0.70. Removing poorly correlated items can enhance internal consistency. After confirming satisfactory reliability, the individual items’ scores are aggregated into a composite score by calculating the mean or sum, representing the overall construct (Gliem & Gliem, 2003).

Relationships Between Variables

To examine the nature, direction, and significance of relationships between constructs such as cohesion, recognition, autonomy, pressure, and fairness, correlation analysis using Pearson’s correlation coefficient is appropriate if data are normally distributed. For non-parametric data, Spearman’s rank correlation offers a robust alternative.

These analyses reveal whether variables are positively or negatively related, the strength of these relationships, and their statistical significance. Further, regression analysis can identify which variables predict satisfaction with the franchisor, providing deeper insights into the dynamics within franchise relationships (Field, 2013).

Conclusion

Effective quantitative analysis of franchisee data involves carefully preparing and examining the dataset, selecting appropriate descriptive and inferential statistical measures, and applying suitable visualization and reliability checks. By systematically following these steps, researchers can uncover meaningful relationships and inform managerial strategies to enhance franchisee satisfaction and control in international settings.

References

• Allen, E., & Seaman, J. (2007). Likert scales and data analysis. Journal of Educational Measurement, 44(3), 193–210.
• DeVellis, R. F. (2016). Scale Development: Theory and Applications. Sage Publications.
• Everitt, B. S., & Hothorn, T. (2011). An Introduction to Applied Multivariate Analysis with R. Springer.
• Field, A. (2013). Discovering Statistics Using IBM SPSS Statistics. Sage Publications.
• Gliem, J. A., & Gliem, R. R. (2003). Calculating, interpreting, and reporting Cronbach’s alpha reliability coefficient. Proceedings of the Midwest Research to Practice Conference in Adult, Continuing, and Community Education.
• Howell, D. C. (2013). Statistical Methods for Psychology. Cengage Learning.
• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric Theory. McGraw-Hill.
• Neuman, W. L. (2014). Social Research Methods: Qualitative and Quantitative Approaches. Pearson.
• Pallant, J. (2016). SPSS Survival Manual. McGraw-Hill Education.
• Tukey, J. W. (1977). Exploratory Data Analysis. Addison-Wesley.