As I continue to refine my dissertation topic, I find myself drawn to the topic of marketing digital transformation (DX). As a Marketing Technology consultant and entrepreneur, I have observed the importance and impact of DX in shaping business outcomes. Unfortunately, the existing literature concerning this topic is not as robust compared to others. Therefore, it is important for me to delve deeper to enrich my understanding and contribute meaningfully to academic literature.
Currently, I work with larger corporations. However, as I expand my services to smaller businesses, it will become important to learn more about how ever evolving digital landscape may impact the revenue and performance outcomes of those smaller businesses. Therefore, some of the research questions I would like to explore is:
RQ1: Does marketing DX significantly influence the revenue for SMBs?
H1: There is not a significant difference in revenue between SMBs that implement marketing DX and those that do not.
Ha: There is a significant difference in revenue between SMBs that implement marketing DX and those that do not.
The appropriate statistical test for this hypothesis would be an independent samples t-test since I will be comparing the means of two independent groups: SMBs with marketing DX and those without (Gerald, 2018).
RQ2: What is the relationship between marketing digital transformation and the overall performance of SMBs?
H2: There is no significant relationship between marketing digital transformation and the overall performance of SMBs.
Ha: There is a significant relationship between marketing DX and the overall performance of SMBs.
The appropriate statistical test for this hypothesis would be a Pearson correlation to help measure the strength and direction of the relationship or association that exists between marketing DX and overall performance of SMBs.
Determining the statistical testing for a research question is contingent on the methodology and framework for the overall research. Choosing the appropriate test is driven by what is attempted to being analyzed, such as determining significance, correlation, differences, or variance. The variables available will also begin to determine which test is used, for example if only one variable exists in determining significance, the one sample t-test would be appropriate whereas with two variables the two-sample t-test would be used. For the hypothesis, it is partially determined by the validation of assumptions once the testing has been satisfied (Wells & Hintze 2007).
Looking at research question I have currently formulated for my qualitative dissertation, one could possibly use a statistical test for measurement. “As a follower does the style or traits of your leader lend to success as team?”. There is an attempt to gauge understanding from a followers’ perspective if their leader is seen as effective in the way they lead. To measure effectiveness, Taherdoost, (2019) suggests a Likert Scale can be used to rate responses in a survey to synthesize results into data points. Likert scales are most likely using ordinal responses to create the data for statistical testing. Once data has been obtained, a Pearson’s Correlation could be used to analyze the relationship between values against performance matrices. Seeing if there is a correlation between the traits and successful performance could begin to identify if one trait is more productive than another. The reasoning for the Pearson’s Correlation is that attempting to find the strength of the relationships between the variables.
Using the following hypothesis: “HO = Team cohesion could be improved by matching leadership styles to followers’”, using an a linear regression to see if there is a relationship between leadership styles and followers who work for leaders with conducive styles. Twomey and Kroll (2008) suggest that linear regression is the best method to find linear relations between coefficients and create associations. They also suggest the visualization shows the relationship and correlation more easily than some other forms of analysis. The independent and dependent variables by again using leadership traits against success, then pairing it with preferred styles of the followers. The intent would be to see if there is a positive linear relationship with a strong correlation that can demonstrated.