Hi everyone,
I was wondering how to get the correlation between one dependent variable and independent variable, Pearson correlation is not an option as there is a follow up visit for each patient (repeated measure), I know mixed model would be an option, but I have a pilot study which means I can have only descriptive stat not inferential stat.
intraclass correlation(ICC) will give me the correlation between those repeated measure not between dependent variable and those repeated measure, am I right? what should I use?
Thanks,
Bahar
If you independent variable is
A) dichotomous (binary or yes/no)
and your dependent variable is
a) also dichtomous, then run the phi coreelation or the tetrachonic one is the underlying measure is nominal)
b) ordinal, the run the Spearman rank-order correation
c) interval, the run the biseral correlation
B) ordinal
and your dependent variable is
a) dichotomous, then ?
b) ordinal, then run the Spearman/Kendal or polychoric correaltion
c) interval, then idem as just above
C) interval
and your dependent variabel is
a) dichotomous, then run the point-besrial correaltion
b) ordinal, then run the polyserial or biserial correaltion
c) interval, then run the pearson correaltion
Thank you for you answer; both dependent and independent var are continuous, I was just wondering, as I have a pilot study can I say:
The analysis of pilot studies should be based on descriptive stat and ideally not involve any inferential stat (i.e., P values). Because a pilot study is conducted to test the feasibility of methods rather than to test hypotheses.
Given the follow-up visits for each patient, Using mixed-effects models without focusing on p-values allows us to get a repeated measure correlation. is that correct?
Y could run a regression analysis and report the R-squared.
R-squared gives a measure of how predictive the regression is and how much variation is explained by the regression. The lowest R-squared is 0 and means that the points are not explained by the regression whereas the highest R-squared is 1 and means that all the points are explained by the regression line
hola: visto el dato repetitivo es cuantitativa, diseño de investigacion longitudinal, es necesario aplicar analis de tendencia, utilizando estadistica descriptiva.
To get the correlation in a repeated measures study:
This approach helps understand the relationship between repeated measures over time.
In a repeated measures study, where the same participants are measured multiple times or under different conditions, you can calculate the correlation between the repeated measures using the following steps:
1. **Prepare the Data**: Organize your data in a format where each row represents a participant, and the columns represent the different measurement times or conditions.
2. **Calculate the Correlation Coefficient**: There are several options to calculate the correlation in a repeated measures design:
a. **Pearson Correlation**: If your data meets the assumptions of normality and linearity, you can calculate the Pearson correlation coefficient between the repeated measures. In this case, you would calculate the correlation between the values at each pair of measurement times or conditions.
b. **Intraclass Correlation Coefficient (ICC)**: If your data has a hierarchical structure, where multiple measurements are nested within each participant, you can use the Intraclass Correlation Coefficient (ICC) to assess the reliability or consistency of the repeated measurements. The ICC takes into account the within-participant and between-participant variability.
c. **Multilevel Modeling**: For more advanced analysis, you can use multilevel models (also known as mixed-effects models) to estimate the correlation between repeated measures while accounting for the hierarchical structure of the data and other covariates.
3. **Interpret the Correlation Results**: Interpret the correlation coefficient (Pearson or ICC) based on the guidelines for the strength of the correlation (e.g., weak, moderate, strong).
It's important to note that the choice of the appropriate correlation method depends on the structure of your data, the assumptions of the statistical tests, and the research questions you are trying to answer. Consulting a statistician or referring to relevant literature in your field can help you determine the most suitable approach for your repeated measures study.
- Badges
- Science topic
- Similar topics
- Biological Science
- Histology
- Sectioning