StackOverflow2013

Note that there are some explanatory texts on larger screens.

plurals

PO
text
Body
copied!<p>Are the two data sets ordered, or not?</p> <p>If ordered, are the indices the same? equally spaced? </p> <p>If the indices are common (temperatures measured on the same days (but different locations), for example, you can regress the first data set against the second, and then test that the slope is equal to 1, and that the intercept is 0.<br> <a href="http://stattrek.com/AP-Statistics-4/Test-Slope.aspx?Tutorial=AP" rel="nofollow noreferrer">http://stattrek.com/AP-Statistics-4/Test-Slope.aspx?Tutorial=AP</a> </p> <p>Otherwise, you can do two regressions, of the y=values against their indices. <a href="http://en.wikipedia.org/wiki/Correlation" rel="nofollow noreferrer">http://en.wikipedia.org/wiki/Correlation</a>. You'd still want to compare slopes and intercepts. </p> <p>====</p> <p>If unordered, I think you want to look at the cumulative distribution functions <a href="http://en.wikipedia.org/wiki/Cumulative_distribution_function" rel="nofollow noreferrer">http://en.wikipedia.org/wiki/Cumulative_distribution_function</a></p> <p>One relevant test is Kolmogorov-Smirnov: <a href="http://en.wikipedia.org/wiki/Kolmogorov-Smirnov_test" rel="nofollow noreferrer">http://en.wikipedia.org/wiki/Kolmogorov-Smirnov_test</a></p> <p>You could also look at</p> <p>Student's t-test, <a href="http://en.wikipedia.org/wiki/Student%27s_t-test" rel="nofollow noreferrer">http://en.wikipedia.org/wiki/Student%27s_t-test</a></p> <p>or a Wilcoxon signed-rank test <a href="http://en.wikipedia.org/wiki/Wilcoxon_signed-rank_test" rel="nofollow noreferrer">http://en.wikipedia.org/wiki/Wilcoxon_signed-rank_test</a></p> <p>to test equality of means between the two samples. </p> <p>And you could test for equality of variances with a Levene test <a href="http://www.itl.nist.gov/div898/handbook/eda/section3/eda35a.htm" rel="nofollow noreferrer">http://www.itl.nist.gov/div898/handbook/eda/section3/eda35a.htm</a></p> <p>Note: it is possible for dissimilar sets of data to have the same mean and variance -- depending on how rigorous you want to be (and how much data you have), you <em>could</em> consider testing for equality of higher moments, as well.</p>

Querying!

Guidance

An individual column

Larger individual text columns get their own page to allow for proper reading.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload