This paper examines the appropriate methods for analyzing business survey data collected using ordinal Likert-scale responses (e.g., "strongly agree" to "strongly disagree"). It explains why nominal, interval, and ratio measurement approaches are unsuitable for this type of data and argues that frequency distribution is the only valid analytical method. Using a 20-respondent employee survey covering 12 questions on corporate leadership, recognition, and job satisfaction, the paper tabulates raw responses, calculates percentages, and ranks questions by favorability. The analysis reveals that while 90.42% of responses were favorable overall, employee recognition and perceived leadership impact scored lowest, suggesting areas for management improvement.
The paper demonstrates methodological justification — the practice of explicitly defending an analytical approach based on data type constraints. By explaining why means, percentiles, and other statistical measures cannot be validly applied to non-numeric ordinal responses, the author shows awareness of measurement theory and its real-world implications for research design.
The paper opens by defining the four response types and situating the survey within the ordinal category. It then addresses why more advanced statistical analysis is inappropriate, before proposing frequency distribution as the correct method. The core findings are presented through three appendices (raw counts, percentages, and ranked results), followed by a concluding interpretation that translates the data into specific management recommendations regarding employee recognition and leadership communication.
There are four general response types for data: nominal, ordinal, interval, and ratio. The most appropriate type depends on the kind of information being gathered. In the survey under analysis, respondents answered each question on the basis of scaled responses such as "strongly agree" or "disagree." These responses, which carry no specific numeric value, are ordinal in nature. The five levels of response are ordered, but there is no precise definition of the differences between categories — the distinctions are entirely subjective from the respondent's point of view.
The responses can be tabulated so that a percentage of answers in each category is calculated. However, no in-depth statistical analysis can be undertaken. Computing a mean, for example, would require numeric values to be assigned to each response. If this had been the researcher's intent, respondents should have been asked for numeric answers. Instead, they were asked for non-numeric answers with ill-defined differences between answer categories. Under these conditions, conventional statistical analysis yields no meaningful results.
Given these constraints, the only viable option for analyzing these responses is within the context of ordinal measurement. A frequency distribution is therefore the only reasonable type of analysis that can be conducted. S. Stevens proposed that ordinal responses could be evaluated using means and percentiles, but even this approach is impossible here because the responses are non-numeric.
The tabulated responses are provided in Appendix A. These results can also be expressed in percentage form for each specific response, as shown in Appendix B. The results can further be ranked by favorability — that is, by the number of favorable responses — to determine where the company scores highest. A secondary sort parameter based on the number of "strongly agree" responses, followed by "agree" responses, produces the ranking shown in Appendix C.
For most questions, the company scored 100% favorable. The only two exceptions were the items concerning recognition for contributions and leadership making changes that are positive for the individual employee. Among the remaining categories, the strongest response came from the question about job satisfaction. Questions about corporate leadership were more mixed. For example, only 8 of 20 respondents strongly agreed that the company is a leader in the industry. Although all respondents agreed that leadership has a clear vision of the future and that the company is a strong competitor, none felt strongly about either statement — indicating a weaker endorsement than for most other questions.
Total surveys completed: 20 | Total survey questions: 12
You’re 37% through this paper. Sign up to read the remaining 2 sections.
Sign Up Now — Instant Access Already a member? Log inAlways verify citation format against your institution’s current style guide requirements.