Data:

                       Trial       1      2      3      4      5      6      7      8       9

       Standard device   450   480   510   500    470   520   475   480   490

       New Device         460   485   505   510    470   525   485   500   510

       Differences(N-S)    10      5    -5     10       0       5    10     20    20 

                      :  8.333        Σx: 75          Σx²:  1,175           N:9 

                        Σx²   –     (Σx)²

       S²d =                          N        =        1,175  -   625       =     68.75

                              N-1                                        8

     SEd =     √S²d/N      =      √68.75/9    =       √7.638889   =  2.76385

      t (calc)    =           đ/SEd   =      8.3333/2.76385    =   3.015

      t (0.02, 8df)   =     3.00

     Therefore reject Ho.

     Summary:  The two devices produced differing results.  Overall, the new device recorded an average

    CO2 level 8.333 ppm higher than the standard device.  The was statistically significant at the 0.02 level. 

    In other words, a difference of this magnitude is not likely (less than 2% likely) to have arisen by chance/random

    variation.  The null hypothesis of "no difference" is rejected.

                             đ: 8.333ppm       t  =  3.015     8 df,     P<0.02