Covariance values greater than 1

[jrhoworth]

I am trying to measure the covariance between subjective well-being (life evaluation) and a number of covariate variables (as shown below), however I keep getting values greater than 1 or lower than -1, what does this mean?


| lifeev~2 age g1
-------------+---------------------------
lifeevalua~2 | 37.0161
age | .931818 2.55303
g1 | .70303 -.080808 .242424
res1 | -2.30071 -.502525 -.250505
ethnicity1 | .413737 .10101 -.016162
employment1 | .200606 -.217172 -.026263
pincome | -.117879 .724747 -.038384
hincome | .092121 .171717 .028283
parentsrship | .270606 .063131 -.024242
personalrs~p | -2.60111 .037879 -.074747
friends | 1.03162 .088384 .153535
siblings | .546667 .232323 -.012121
placeofres~1 | .006263 .136364 .008081
education1 | 2.51838 .969697 .048485
deaths1 | .140909 -.04798 .010101

[David Hoaglin]

Correlations can't be greater than +1 or less than -1, but covariances are not subject to any such restriction. Recall that, for the usual product-moment (Pearson) correlation coefficient of X and Y, the definition is corr(X,Y) = cov(X,Y)/[SD(X)SD(Y)].