I would appreciate help in checking whether my attempt to decide whether my correlations are significant is correct – I suspect not, and would welcome a correct set of functions.
After computing the correlation of two arrays, for example: JL50: CORREL($QW$5:$QW$46,JL$5:JL$46) = 0.4084,
I then compute the t value: J51 =JL50*SQRT(JL47-2)/SQRT(1-JL50^2), where JL47 = n = 31, giving 2.4094
From this, I then compute the significance of the correlation using: J52 =T.DIST(JL51,JL47-2,2) = 0.9887
Finally, I test if this is significant by =IF(ABS(JL$52)<$D$49 ,””, SLOPE(JL$5:JL$46,$QW$5:$QW$46)) where $D$49 = 5%
which gives a blank, yet I feel that with the computed correlation and t value, this ought to indicate my correlation is significant. The above seems to work fine for negative correlations.
After computing the correlation of two arrays, for example: JL50: CORREL($QW$5:$QW$46,JL$5:JL$46) = 0.4084,
I then compute the t value: J51 =JL50*SQRT(JL47-2)/SQRT(1-JL50^2), where JL47 = n = 31, giving 2.4094
From this, I then compute the significance of the correlation using: J52 =T.DIST(JL51,JL47-2,2) = 0.9887
Finally, I test if this is significant by =IF(ABS(JL$52)<$D$49 ,””, SLOPE(JL$5:JL$46,$QW$5:$QW$46)) where $D$49 = 5%
which gives a blank, yet I feel that with the computed correlation and t value, this ought to indicate my correlation is significant. The above seems to work fine for negative correlations.