Data analysis - Hypothetical

This segment focusses on different potential analytical techniques the researchers may use to investigate Verizon Wireless Store's problem of customer satisfaction. Until the study is carried out, however, the data should undergo essential procedures, including code and cleaning. Coding and cleaning operations are essential to maintaining a reliable and ultimately dependable data processing by the researcher. The handling of missed cases, inability to respond and invalid findings is part of the cleaning process. The outlined data collection process entails several variables including the customer’s satisfaction that is the dependent variable of interest. In this case, the study explores how various approaches to data analysis can be operationalized in the current case study of Verizon Wireless Communications.

The study is both quantitative and qualitative or in other words; it is a mixed-method study design. Therefore, there a myriad of statistical analysis procedures aligned with both qualitative and quantitative data that one can use to analyze the data.  In fact, the analysis can apply both parametric and non-parametric methods of data analysis. Non-parametric methods are described as those with assumption-free distributions while parametric methods are described to follow Gaussian distributions which are characterized by assumptions. Specifically, for the customer satisfaction problem at Verizon Wireless some of the data analyses procedures the study might utilize can be broadly categorized as descriptive, inferential, and trend analysis which includes probability, linear regression, and time series.

Usefiul info: GrabMyEssay is your best college application essay writing service!

Descriptive statistics

  The descriptive statistics are fundamental features in any given study. The importance of using descriptive statistics to make it easy for non-statisticians understand the data collected by providing summaries of the sample (Cortinhas, 2012). In most studies, descriptive statistics are combined to with graphics analysis. This combination forms the basis for further investigation. In the current case, data will be collected from several variables including measures of customer satisfaction such by asking the customer’s to rate the level of satisfaction with Verizon Wireless Stores between 1-100 where is a 100 is the highest satisfaction score. The study will also assess other factors such as billing costs, network performance and reliability, communication, and customer service. The descriptive statistics needed to summarize this data include frequencies or counts, measures of central tendency and measures of variability.

In most specific terms, frequency counts will help summarize most of the qualitative data variables such as the customer service and competence of employees. The two variables as explained in the survey questionnaire are categorical, ordinal variables; therefore, by counting each category, it will be easier to explain how the customer rates the level of customer service and employees competence for Verizon Wireless Stores. The frequencies of these categorical variables will help create visual diagrams such as bar chart or pie chart for purposes of establishing the suitable impression of the raw data. Additionally, the frequencies will be used to explore the distribution of various quantitative variables such billing costs, rating on network performance and reliability, and customer satisfaction rating by creating histograms. Measures of central tendency such as mean, mode, and median will also provide analysis for the quantitative variables. In particular, these measures will help determine to summarize customer satisfaction ratings, network performance and reliability rating, and billing costs. Finally, the measures of variability including the variance and standard deviation will be used to estimate the variation of observation in the customer satisfaction ratings, network performance and reliability rating, and billing costs.

Ask GrabMyEssay "I need help with my research paper" and let experienced writers assist you with writing tasks of any complexity! 

Inferential statistics

The purpose of the inferential statistics extend beyond summarize data alone. Inferential statistics are helpful in making substantiated conclusions about the data (Sekaran & Bougie, 2013). Most inferential statistics are used to compared groups or determine dependencies. This type of statistics includes the chi-square statistic, t-statistic, z-statistic, F-statistic, and others that are related to the non-parametric test. However, not all these statistics will be included in a study; and therefore, it is the work of the researcher to determine those that best fit the study hypotheses. The chi-square test applies mostly when testing for association between categorical variables; therefore, it is a non-parametric test.

In case of customer satisfaction study for Verizon wireless, the researcher will use chi-square test of association to determine whether there is an association between a  demographic variable such as region and customer service. The aim is to test the hypothesis that customer service rating (on a 5-point Likert scale) is not associated with region (location of Verizon wireless store) from which the customer accesses the services. In other case, t-test statistics can help determine whether the customer satisfaction rating varies among young persons (40 years and below) and older persons (above 40 years). Using the one-way ANOVA, the study will calculate f-statistic to compare whether the level of customer satisfaction with Verizon wireless stores across the five residential regions (South, North, West, East, and Central).

Useful info: Where I can hire an expert to do my essay online?

Probability in the analysis

Probability is very helpful in ensuring the tests are used to making meaningful conclusions. Every inferential statistic tests a specific hypothesis, but it is not complete until the researcher determines the probability of making type I error or instead of the alpha level (Banerjee, Chitnis, Jadhav, Bhawalkar, & Chaudhury, 2009). The alpha level is a probabilistic value that provides for the significance of test hypotheses. In the current study, the entire necessary hypothesis will be tested at the 0.05 level of significance. The probability will also assist in estimate confidence intervals for the true mean value of customer satisfaction ratings with Verizon wireless stores.

Linear regression analysis

The regression analysis is an important tool for studying causation between the dependent variable and possible predicting factors (Montgomery, Peck, & Vining, 2012). In particular, the linear regression analysis is used to study causal relationship between variables, which are thought to have linear relationships. However, variables do not just enter regression without satisfying various underlying assumption. These assumptions include linearity, the presence of outliers, normality of residual errors, and independence of observations. The assumption of linearity is determined by checking for linear relationship using scatterplots. The presence of outliers is also assessed using a scatterplot. In the current case study, the researcher will conduct a multiple linear regression analysis whether the customer satisfaction ratings is response variable and the billing costs and network performance and reliability rating (1-100) as the explanatory factors.

Time series analysis

Time series analysis is an appropriate technique when accounting for data observed over time that may possess some internal structure such as trend. Though the current study is largely cross-sectional, it is important to understand whether then Verizon wireless stores are working towards providing better services for their customers. In this case, the aim is to explore whether a total number of customer complaints regarding customer service and network performance over the next 30 days will change. In other words, a time series equation will be constructed to determine whether there is any trend in the number of complaints over time.

Read also: Masters essay writing service is a must-have for students who want to have high grades but don't have enough time.

References

Banerjee, A., Chitnis, U., Jadhav, S., Bhawalkar, J., & Chaudhury, S. (2009). Hypothesis testing, type I and type II errors. Industrial Psychiatry Journal, 18(2), 127. doi:10.4103/0972-6748.62274

Cortinhas, D. C. (2012). Business Statistics for Contemporary Decision Making - 1st European Edition: First European Edition. Hoboken, USA: Wiley.

Montgomery, D. C., Peck, E. A., & Vining, G. G. (2012). Introduction to linear regression analysis (5th ed.). Hoboken, NJ: John Wiley & Sons, Inc.

Sekaran, U., & Bougie, R. (2013). Research methods for business: A skill-building approach. Chichester, West Sussex: Wiley.