I am asking one group of people some survey questions using a 5 point Likert scale (strongly disagree to strongly agree). I was planning on using descriptive statistics such as mode, median and frequency distributions. Should I also be using any other statistical analyses? I know that you can use Mann-Whitney U test but it looks like these would only be used if I was comparing the responses of 2 different groups?
Rebecca Lee It entirely depends on what you want to know from these data...What prompted you to collect the data in the first place?
When analyzing data collected using a Likert scale, the choice of statistical analysis largely depends on the specific goals of your study and the nature of your data. Here are some common statistical methods used for Likert scale data:
1. Descriptive Statistics: These include calculating the mean, median, and mode to understand central tendencies, as well as measures of variability such as range and standard deviation. Since Likert scale data are ordinal, the median and mode are often more appropriate measures of central tendency than the mean.
2. Frequency Distribution: Counting and reporting the frequency of each response can provide insights into the distribution of opinions or attitudes within your dataset.
3. Cross-tabulation: This involves examining the relationship between two categorical variables (which can include Likert scale items). Chi-square tests can be used to determine if differences in responses are statistically significant.
4.Correlation Analysis:
Spearman's Rank Correlati: Suitable for ordinal data, this non-parametric test assesses how well the relationship between two variables can be described using a monotonic function.
Kendall’s Tau: Another non-parametric correlation measure, suitable for small samples or data with many tied ranks.
5. Factor Analysis:
Exploratory Factor Analysis (EFA): Often used to identify underlying variables, or factors, that explain the pattern of correlations within a set of observed variables (Likert items).
Confirmatory Factor Analysis (CFA): Used to test the hypothesis that a relationship between observed variables and their underlying latent constructs exists.
6. Reliability Testing:
Cronbach’s Alpha: Used to estimate the reliability, or internal consistency, of a Likert scale. This is especially important if the scale includes multiple items meant to measure a single construct.
7. Regression Analysis:
Ordinal Logistic Regression: Appropriate when the dependent variable is ordinal (such as Likert scale responses), this method can be used to model the relationship between the dependent variable and one or more independent variables.
8. Non-parametric Test (for smaller sample sizes or when data do not meet parametric assumptions):
Mann-Whitney U Test and Wilcoxon Signed-Rank Test: Used for comparing two independent or paired samples, respectively.
Kruskal-Wallis H Test: An extension of the Mann-Whitney U Test for more than two groups.
Choosing the right statistical analysis will depend on whether your Likert scale data is considered ordinal or interval, the number of Likert items, the distribution of your data, and your research hypotheses. It's also important to consider the size of your sample and whether your data meet the assumptions required for parametric tests. Non-parametric tests are a safer choice when these assumptions are not met.
- Badges
- Science topic