Case Chapter 15: Quantitative Data Analysis The Subway Famil

Case Chapter 15 Quantitative Data Analysisthe Subway Family

In 1965, Fred DeLuca and Peter Buck opened the first Subway sandwich shop in Bridgeport, Connecticut, with a goal of expanding to 32 shops in ten years. By 1974, they operated 16 shops and began franchising to accelerate growth, leading to Subway becoming one of the world's fastest-growing franchises. The regional office for European franchises is located in Amsterdam, where Fin Green, a Senior Manager, faces the challenge of maintaining control over franchisees without stifling their entrepreneurial spirit. To address this, Fin asked Bart Veldkamp, a Master's student, to investigate franchisee satisfaction using a conceptual model based on research by Strutton et al. (1993). Bart developed a questionnaire measuring six independent variables—Cohesion, Recognition, Autonomy, Pressure, Fairness—and the dependent variable, Satisfaction with the Franchisor. The questionnaire includes Likert-scale items from 1 (Totally disagree) to 7 (Totally agree). It also gathers demographic data such as age, gender, education, and years as a franchisee. Bart has collected data and needs to analyze it appropriately, including preliminary steps, measures of central tendency and dispersion, handling of data inconsistencies, visual data exploration, reliability testing, and correlation analysis. This process aims to understand the relationships among the variables and assess franchisee satisfaction comprehensively.

Paper For Above instruction

Analyzing data collected from franchisees through surveys requires a systematic approach to ensure accurate and meaningful insights. The preliminary steps begin with data cleaning and organization, which involve checking for missing values, outliers, and inconsistencies that could distort analysis. Data entry should be verified for accuracy, and any incomplete or invalid responses should be addressed, either by imputation or exclusion. Once clean, coding categorical responses and assigning numerical values to Likert scales are essential for quantitative analysis. Descriptive statistics then provide initial insights into the data, guiding further analysis.

To summarize the central tendency and variability of the questionnaire items, appropriate statistical measures are used. For individual items representing variables such as Cohesion, Recognition, Autonomy, Pressure, and Fairness, the mean and median serve as central tendency indicators, illustrating the average response, while standard deviation and range reveal data dispersion. These measures help assess the general sentiment and consistency among franchisees for each construct. For demographic variables like age and years as a franchisee, measures such as mean and frequency distributions are suitable for understanding sample characteristics. Using these measures complements the initial data overview, informing subsequent inferential analyses.

Regarding the measurement of age and years as a franchisee, potential problems include the categorization scheme. If age and experience are collected in broad categories (e.g., "55"), data granularity is limited, reducing precision in analysis. Such categorical coding can obscure variations within groups and hinder the detection of linear relationships. A better approach might be to gather exact ages and tenure durations, allowing for continuous data analysis and more nuanced statistical testing, such as correlation or regression analysis.

Visual data exploration enhances understanding of variable distributions and outliers. Histograms display the frequency distribution of continuous variables, enabling detection of skewness, modality, and outliers. Box-and-whisker plots summarize data symmetry, spread, and potential outliers with clear visual cues. While pie charts are useful for categorical distributions, histograms and box plots are more informative for interval or ratio data, especially when examining scale items. Providing these plots for all items is advisable because they reveal distributional properties essential for choosing appropriate statistical tests and for identifying data issues like skewness or outliers.

Reliability testing of multi-item scales ensures the consistency of responses across items intended to measure the same construct. Cronbach’s alpha is a common statistic used, with values above 0.70 generally indicating acceptable internal consistency. Reverse scoring may be necessary when some items are negatively worded; failing to reverse their scores would lead to inconsistent responses that artificially lower alpha. For example, items like "I feel like I never have a day off" or "too many people get burned out" might be negatively worded and require score reversal to align with positively worded items, ensuring composite scores accurately reflect the underlying construct.

Concerning Cronbach’s alpha for cohesion, the provided table indicates a value of 0.804 with five items. Since alpha exceeds the 0.70 threshold, the scale demonstrates good internal consistency. Item analysis reveals that removing items with low corrected item-total correlations or those that increase alpha if deleted can improve the scale’s reliability. In this case, item cohesion2 has a relatively low correlation, suggesting it might be less consistent with others. Once the scale's reliability is confirmed, the individual item scores are summed or averaged to create a composite measure of cohesion, facilitating analysis and interpretation.

To examine relationships between variables such as Cohesion, Recognition, Autonomy, Pressure, and Fairness, correlation analysis (e.g., Pearson’s r) is suitable. These bivariate analyses quantify the strength and direction of relationships, indicating whether variables move together positively or negatively. Additionally, significance testing (e.g., p-value) assesses whether observed correlations are statistically meaningful. Such analysis helps determine how factors like cohesion or fairness influence satisfaction with the franchisor, guiding strategic decisions aimed at improving franchisee relations.

References

• Field, A. (2013). Discovering Statistics Using IBM SPSS Statistics. Sage Publications.
• Carmines, E. G., & Zeller, R. A. (1979). Reliability and Validity Assessment. Sage Publications.
• DeVellis, R. F. (2016). Scale Development: Theory and Applications. Sage Publications.
• Hibbard, J. H. (2002). Making Sense of Cronbach’s Alpha. Journal of Nursing Education, 41(10), 469-472.
• Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate Data Analysis. Pearson.
• Strutton, D., Pelton, L., & Lumpkin, J. (1993). The influence of psychological climate on conflict resolution strategies in franchise relationships. Journal of the Academy of Marketing Science, 21(4), 23-31.
• Tabachnick, B. G., & Fidell, L. S. (2013). Using Multivariate Statistics. Pearson.
• Anton, M., & Neuman, W. (2014). Survey Data Collection and Analysis. Oxford University Press.
• Blumberg, B., Cooper, D. R., & Schindler, P. S. (2014). Business Research Methods. McGraw-Hill Education.
• George, D., & Mallery, P. (2010). SPSS for Windows Step by Step: A Simple Guide and Reference. Pearson.