Please i need help , i am working on salt stress and the effect of humic acid application . i have 4 variety of plant and 4 traitement (0 g/l ,2g/l, 4 g/l and 6g/l). every traitement divded into : 3 groups . for exemple 0 g/l without humic acid application ; 0g/l with humic acid application H1 and 0 g/l with humic acid application H2. i need to know whish test should i make with spss to compare all these traitement and to show how significant the traitement with humic acid was to alleviate the salt stress in 0 ,2,4 and 6 concentration of salt ?
You can run a full factorial design (general linear model), If the interaction between factors results significant, then the interaction graph will be very helpful to compare the treatments that you mentioned.
otherwise, you can use of ANOVA.
If I understood correctly, your variables are nominal. If so crosstabs under descriptive statistics in SPSS would be useful (please remeber to tick Chi-square in Statistics to obtain info whether the differences are statistically significant). Good luck.
THANKS FOR YOUR REPONSES . i want to know how to compare all traitement (interactions between traitement) for exemple if traitement with humic acid H1 with application 2 g/l of salt was better than the traitement with humic acid H2 with application of 4 g/l of salt ? I KNOW HOW to compare traitement . for exemple traitement with humic acid H1 IS better than traitement H2. BUT I SHOULD KNOW in which concentration it is better ? in 0 or 2 or 4 or 6 of salt .
dear professor Gonzalez , can you explain me by step the procedure . i should be grateful
Dear professor , this attached paper can explain a little . Thank you for your time .
You can run a full factorial design (general linear model), If the interaction between factors results significant, then the interaction graph will be very helpful to compare the treatments that you mentioned.
otherwise, you can use of ANOVA.
It is necessary to know what your study design is. Are you doing repeated measurements or taking single measurements with the different concentrations of humic acid? If you are doing repeated measurements, then your variables are correlated therefore you need repeated-measure ANOVA but if your measurements are single for each concentration then use 2-way ANOVA. these two methods are based on the assumption that you are data are normally distributed. if they are not, then use Kruskal wallis for your analysis. All the best.
I think that your data must test for normality before you do ANOVA , in case of normal distribution , you can analyze your data with full factorial , then you can do the comparisons between means by Duncan multiple range test
Good Luck
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis