my research is about the nutritional status of hemodialysis patients.
I have data from patients who use a snack and who do not use a snack.
Both groups repeated the same test in December 2017, June 2018, December 2018 and May 2019.
I investigate whether patients who eat a snack have a better nutritional status. I would like to calculate the significance between T0, T1, T2 and T3. But also mutually T1 to T2; T2 to T3; T1 to T3.
Which SPSS test is most suitable for this? Or do I have to combine tests?
Dear Petra Krul
SPSS is not a statistical test; is a pre-programmed statistical tool or package that contains databases and various statistical tests. It is managed on a computer and the latest version is compatible with Windows. There are also other statistical packages that you can use, such as EPISTAT and STATA.
In the case of the experiment you are designing or planning to carry out, you must first identify which variables are qualitative and which others are of a quantitative nature; once you have done this, the next step is to prepare the database with the results of your research, considering the zero time and the different cutting times (T0, T1, T2, T3 etc).
From the beginning it can be seen that its design will probably be of the factorial type of one or two tracks and at least of the 2x3 plate.
With the data already ordered in the matrices, both qualitative and quantitative data (separately); The Variance Analysis will be applied for 1 or 2 routes (AVAR 1, AVAR 2) for quantitative variables. Likewise, the qualitative data will be analyzed with Chi Square (X2) for nominal variables and Kruskal Wallis or Freedman test for ordinal variables.
If another type of analysis and statistical tests are required, we can discuss it throughout this network.
regards Jose Luis
https://www.stat.ncsu.edu/people/davidian/courses/st790/notes.html
Is the best source around on longitudinal models which is what you have. Pay particular attention to plots of the 2 groups versus time to see if they look different. Since you don't have many repeats the chapter 1 methods may suffice. In any case what you need is here Best, David Booth
What is your intervention? Is it having a certain snack available? and if so how were people allocated to this versus a control group?
Dear Petra Krul
SPSS is not a statistical test; is a pre-programmed statistical tool or package that contains databases and various statistical tests. It is managed on a computer and the latest version is compatible with Windows. There are also other statistical packages that you can use, such as EPISTAT and STATA.
In the case of the experiment you are designing or planning to carry out, you must first identify which variables are qualitative and which others are of a quantitative nature; once you have done this, the next step is to prepare the database with the results of your research, considering the zero time and the different cutting times (T0, T1, T2, T3 etc).
From the beginning it can be seen that its design will probably be of the factorial type of one or two tracks and at least of the 2x3 plate.
With the data already ordered in the matrices, both qualitative and quantitative data (separately); The Variance Analysis will be applied for 1 or 2 routes (AVAR 1, AVAR 2) for quantitative variables. Likewise, the qualitative data will be analyzed with Chi Square (X2) for nominal variables and Kruskal Wallis or Freedman test for ordinal variables.
If another type of analysis and statistical tests are required, we can discuss it throughout this network.
regards Jose Luis
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS