Look for a mistake. In real biological data it is basically impossible to get an R2 of 1. Go back to the original data on the original data sheets. Start there and check all the numbers. Check the transformations (if any). Check the steps that tell the statistics program how to run the analysis. If everything looks right then try some fake data. Data where you know the answer because that is how you got the numbers. Alternatively, use a greatly reduced data set. Then calculate the regression manually to see that the program and you agree.

I notice that you have no values for F-value and P-value. This is another indication that something is wrong. I note that the standard error for the coefficients is zero in al cases. Might one variable be a constant?Unfortunately, I don't know SPSS well enough to make further guesses.

Look for a mistake. In real biological data it is basically impossible to get an R2 of 1. Go back to the original data on the original data sheets. Start there and check all the numbers. Check the transformations (if any). Check the steps that tell the statistics program how to run the analysis. If everything looks right then try some fake data. Data where you know the answer because that is how you got the numbers. Alternatively, use a greatly reduced data set. Then calculate the regression manually to see that the program and you agree.

I notice that you have no values for F-value and P-value. This is another indication that something is wrong. I note that the standard error for the coefficients is zero in al cases. Might one variable be a constant?Unfortunately, I don't know SPSS well enough to make further guesses.

Dear Johannes,

R-square tell us how much variation in the dependent variable is accounted for by the regression model, the adjusted value tells us how much variance in the dependent variable would be accounted for if the model had been derived from the population from which the sample was taken. Specifically, it reflects the goodness of fit of the model to the population taking into account the sample size and the number of predictors used.

A data-set should always be explored to see if it meets the assumptions of the statistical methods applied. The multivariate data analyses we are intending assume normality, linearity and absence of multicollinearity .

Normality refers to the shape of the data distribution for an individual variable and its correspondence to the normal distribution. In this study, the assumptions of normality were examined by looking at histograms of the data, and by checking skewness and kurtosis. The distribution is considered normal when it is bell shaped and values of skewness and kurtosis are close to zero.

The linearity of the relationship between the dependent and independent variables represents the way changes in the dependent variable are associated with the independent variables, namely, that there is a straight-line relationship between the independent variables and dependent variable. This assumption is essential as regression analysis only tests for a linear relationship between the independent variables and dependent variable. Pearson correlation can capture the linear association between variables.

Multicollinearity is the existence of a strong linear relationship among variables, and prevents the effect of each variable being identified. Many Scholars recommend examining the variable inflation factor (VIF) and tolerance level (TOL) as a tool for of multicollinearity diagnostics. VIF represents the increase in variance that exists due to collinearities and interrelationships among the variables. VIFs larger than 10 indicate strong multicollinearity and as a rule of thumb VIFs should be less than 0.1

According to your analysis, An R-square=1 indicates perfect fit. That is, you've explained all of the variance that there is to explain. you can always get R-square=1 if you have a number of predicting variables equal to the number of observations, or if you've estimated an intercept the number of observations . But in response to your general question, You need to check the regressions assumptions mentioned above.

Best Wishes,

Dear Johannes,

try with One Way Anova by Vassarstats. its a free software which compares the means of the data set and says about the significance and the results are self explanatory. also u can try graph pads t test to arrive at precise conclusions. u can send the results to me so that i can analyze and send it to you. i use SPSS also i am familiar with the software

Hi Johannes,

You used four degrees of freedom to fit the regression model. But there are no degrees of freedom for your errors. Clearly this is wrong because it indicates that you do not have any observations in your regression model. Timothy is right. Do check your dataset again.

Haitham makes a good point that my answer missed. You have 5 samples from plot A, and 13 from plot B. The first variable in the ANOVA model needs 4 degrees of freedom. Replication in plot A is insufficient for this model. You could try simpler (univariate) statistical models. You could try collecting more data, possibly next field season.

Sometimes bad things happen to good experiments. However, even if the intent was to have 13 replicates from plot A and 13 from plot B, this would not be sufficient replication to support a model with 18 independent variables.

Thanks for all answers.

Now, I will check my dataset and I will use BLUE (regression) method to analyze my data.

Maybe, I will try used PCA for my statistical analysis.

PCA is an interesting choice, but you still only have 5 replicates in soil A. This will limit the number of variables you can enter into the model. You now have a problem of how to choose which variables to enter.