Questions tagged [normality-assumption]
Many statistical methods assume data or a model's residuals are normally distributed. Use this tag for questions about the assumption & testing of normality, or about normality as a *property*. Use [normal-distribution] for questions about the normal distribution per se.
1,084 questions
2 votes
1 answer
64 views
Paired t-test vs. Wilcoxon signed-rank test
I have a paired dataset with n = 24. I checked the normality of the difference scores using the Shapiro–Wilk test, and p value is =0.043 I ran both: Paired t-test Wilcoxon signed-rank test Both tests ...
1 vote
1 answer
84 views
Should I log-transform individual timepoint measurements or the absolute change score when my outcome is volume (cm³)?
I have a longitudinal dataset where my outcome is white matter hyperintensity (WMH) volume measured in cubic centimeters (cm³) from brain MRI, collected at baseline and follow-up for each participant. ...
- 425
9 votes
4 answers
761 views
Is a normality test always performed on errors and not on raw data?
This question is mainly because I see many tutorials, whether on YouTube videos or on websites, forums, and the like, where when using software like SPSS, they always apply normality tests directly to ...
0 votes
0 answers
39 views
Multilevel analysis assumptions violation and singular warning
I am working on a project with 3 measures of cortisol and 3 measures of infant crying, thus I assess with multilevel if there is an association between the two. I have quite some missing data in my ...
- 1
20 votes
5 answers
1k views
Why assume normal errors in regression?
The errors in linear regression are often assumed to be normal. My understanding is that this is because of the following reasons, if there are more please feel free to let me know: Mathematical ...
- 675
9 votes
5 answers
1k views
Are p-values still useful even though errors are not normal?
Recently, I've been working with a coworker in estimating some multiple linear regressions. His first approach was to show me the output of summary(lm) of R, and he ...
- 171
0 votes
0 answers
60 views
Normality of residuals in aov() can be assumed, but test for homoscedacity fail, do I have to recheck normality for each group applying oneway.test()? [duplicate]
Assume the following situation: Assume you ran a Fisher Oneway-ANOVA (aov() in R) and found that the residuals are normally distributed based on the Shapiro-Wilk ...
1 vote
1 answer
95 views
Bootstrapping - Interpretation and transformation
I am currently trying bootstrapping. My dependent variable is continuous (Change in ROA). My independent variables are GDP growth and inflation rate (both continuous). I also include three control ...
- 11
1 vote
0 answers
115 views
Using robust linear mixed effect model instead of linear mixed effect model
I have one quantitative dependent variable and 2 within-subject factors. I planned to an analysis of this data in R using the lmer command to model my data and then compare models with more and more ...
- 11
1 vote
0 answers
55 views
Confidence Interval for Population with Disproportionate Stratified Random Sample
I am reviewing data from a stratified random sample to describe the proportion of the total population exhibiting Trait A ($p_A$). Due to time/resource constraints, I conducted a disproportionate SRS ...
- 11
0 votes
0 answers
87 views
Shapiro-Wilk and [duplicate]
When testing for normal distribution of given numbers, (Shapiro-Wilk, Kolmogorov-Smirnov, Z-values of symmetry and kurtosis) and graphical methods (Q-Q diagram, box plot, histogram) are used. ...
- 121
3 votes
1 answer
142 views
Expected Value and Variance of Skewness/Kurtosis Estimator for the Difference of Normal Random Variables
As known, sample skewness ($g_1$) and kurtosis ($g_2$) can be calculated as follows: $$ g_1 = \frac{m_3}{m_2^{3/2}} = \frac{\tfrac{1}{n} \sum_{i=1}^n (x_i-\bar{x})^3}{\left[\tfrac{1}{n} \sum_{i=1}^n \...
- 465
6 votes
3 answers
679 views
Which statistics or statistical procedures would show the effects of different types of physical exercises on some hormones?
I am an academician and had an experimental study on the effects of different types of physical exercises on some hormones. I set up three groups: high intensity exercise, low intensity exercise, and ...
0 votes
1 answer
85 views
Should I use mean or median as measure of central tendency?
I am comparing a dataset for 2 biomarkers in a sample size of 30, before and after intervention. Some data is normally distributed and some aren't. I have used appropriate tests for the same. Can I ...
- 1
5 votes
1 answer
282 views
Moderation analysis assumption: univariate outliers after centering
I am conducting a moderation analysis for my thesis and am performing assumption testing. I found a few univariate outliers and transformed any scores that were z-score of > (-)3.29. I then ...
- 51