This is a step-by-step guide to doing a correlation or regression analysis using the Acastat software.
First enter your data into the spreadsheet, making sure you have labelled each variable at the head of each column. If you don't know how to label columns, go here.
To begin a regression or correlation test, select the statistical procedures tab.This opens up a page as below:
Choose your dependent and independent variables.
Select simple regression
Click on the blue Run arrow at the top right of the screen.
Observing, Copying and Pasting Results
After pressing the Run button, the Statistical Procedures page should now show a scatter graph in the bottom right with points on it.
Double-click the graph
This should take you to an output screen showing the scattergraph together with the statistical results of the regreesiion analysis:
You can copy the scattergraph and paste it into a word processor document. (The software gives the graph a poor title, so you'll have to add another one using your word processor).
You can print out your statistical results in order to properly examine them.
What do the Results Actually Mean?
- The Pearson Correlation Coefficient tells you that the 2 variables are HIGHLY CORRELATED. (A correlation coefficient of 1 means a perfect correlation, 0 means no correlation at all).
- The P-value of 0.0046 means your results are HIGHLY STATISTICALLY SIGNIFICANT. (Remember, at A-level we accept results as being statistically significant if P<0.05 and think results are highly significant if P<0.01 as this means that there is a less than one chance in 100 that your findings are due to luck).
- The R-squared result of 0.47 means that 47% of the variation in your dependent variable (% cover in this example) can be explained by changes in the independent variable (pH).
How Do I Lay Out my Results in my Write-Up?
Here's an example:
Soil pH and the %cover of Marram Grass
The relationship between the pH of the soil and the %cover of Marram grass found in random quadrats within a sand dune is illustrated in the graph below. I have added a line of best fit.
The results of a regression analysis are shown in the table on the right.
 |
|
From the graph you can see that as the soil pH increases, the percentage cover of Marram Grass increases. The two are positively correlated.The Pearson Correlation Coefficient of 0.69 shows that the two variables are highly correlated.
The R-squared value of 0.47 shows us that soil pH is an important variable influencing the %cover of Marram grass. It means that 47% of the variation in %cover can be explained by changes in the pH.
The relationship is highly statistically significant, woth P<0.01. This means that there is a less than 1 in 100 chance that the relationship is due to chance alone.
...and in your conclusion you would explain the BIOLOGICAL MEANING and REASONS for your results.
What is your results are NOT statistically significant?
Provide a graph
Explain the meaning of your statistics
Describe the relationship that your graph shows even though it is not statistically significant.