Share
Explore BrainMass

# Testing correlation for signficance.

In both cases, give a number of data points, n, and a correlation coefficient, r. Draw a number of data points, n, and a correlation coefficient, r. Tell whether the
correlation is significant at 0.05 or 0.01 level, explain.

n=50 points, r=0.35.

n=20 points, r=0.58.

#### Solution Preview

OK, now that you've chosen your sample sizes and r values, you should start with drawing them. For the first one, r=.35, let's start with imagining what the slope of this would look like. Remember that when r=0, your slope=0; that is, your line of best fit (the one that fits the data points) goes horizontally straight across. When r=1, your line goes on a perfect diagonal from the bottom left-hand corner of your graph to the top right-hand one. When r=.50, your line would fall half-way between the perfect diagonal we see with r=1, and the flat line we see with r=0. So now, imagine what r=.35 should look like - it should be on a slope just a little less steep than that for r=.50. Once you draw that, you should fill in 50 points that surround that line - half of them should be above it, and half below ...

#### Solution Summary

Interprets whether 2 correlation coeffiecients indicate significant correlations.

\$2.19