Hello,
I am wanting to compare data from my study to a previous study completed in my workplace. They are both a growth study of a buffer over 28days in which data is collected as cfu/ml at each timepoint. As part of my apprenticeship end point assessment i must statistically analyze the data but i am unsure of what would be the best tool to use. I have attached a example of my data below.
First, some questions that must be answered:
1) About your data: Does this table contain all the records? From my understanding, you measured the growth for one week straight (0-7), then once per week (7,14,21,28). This missing daily data might mess with the analysis.
2) Again about the two variables? What do you mean by growth? Is it the difference in cfu/ml from the current timepoint and the previous timepoint?
Former study seems to have no growth by the end (all 0's), while the repeat study shows big growth at the end (710).
3) Finally, what is your question?
If you want to know whether the differences in growth between the original and repeat study are statistically significant, then you should pick either ANOVA or Kruskall-Wallis tests. These will tell you if the growth is different.
If you want to find the level of congruence between them, then you should use Pearson or Spearman correlation to find possible relations.
I would go for Kruskall-Wallis, as it is non-parametric!
4) If you want to model the growth over time, you could create a linear model, which will get the time as independed variable and calculate the depended variable of growth. There are other similar models, or variable transformations to find what you need and fits to your data!
I hope that I helped!
If you are familiar with coding, you could perform these tasks in R or Python! Both are free and you will find plenty of tutorials!
If not, there are other ways, programs like SPSS, minitab or Stata, but as far as I know they are monetized and you must buy a license to use them!
Maybe you could do this on MS Excel, but I'm not 100% sure about that!
Panos Petsas thank you so much for your help!
In regards to your questions:
1. Yes that is all the data I have, and the timepoints you have mentioned are correct.
2. In terms of growth I am referring to microbial growth i.e the number of colonies present at that time not the comparison to the last timepoint.
3. Ideally I would like to see if the differences are significant. I will give the ANOVA and Krushall Wallis a try. If i complete this in excel, can i select all the data for each study as the two data sets, or do i need to consider the timepoints too?
Thanks again!
Chloe Paige Mason Happy to help!
Im ok with 1. and 2.
For 3: I haven't tried this in MS excel, but I think you should select the two columns with the growth data ("original study" and "repeat study"). Not the column with the timepoints!
I googled for this and found this guide for KW in excel
https://www.statology.org/kruskal-wallis-test-excel/
this could help you to do the analysis!
Chloe Paige Mason
And another guide for KW in excel.
This one explains how to get the p-value, with the chi-square distribution, H and degrees of freedom!
https://www.real-statistics.com/one-way-analysis-of-variance-anova/kruskal-wallis-test/
Hello
Tring with excel
Ph.D. Ingrid del Valle García Carreño
You should make a comprehensive analysis of your case study, as the criteria for selecting the statistical test to be used, with the support of heuristics, include the following:
- The research level.
- The statistical objective.
- The type of variable.
- The type of research.
- The research design.
- The behaviour of the data.
- Badges
- Science method