Using the above data check to see if the average cholesterol determination was statistically the same or not for two labs. alpha= 0.05. Use a independent design.
My problem is that I got negative Tcalc. Since it cannot be negative I'm not able to finish step 5, p value. I'm not sure if I did this wrong, or is there a different way to get p value, or should I just ignore negative sign and use value of 1.5621 to finish the last step. Please help. Thank you.
Based on your experiment, I am assuming that you are using the Unpaired Student-t test.
Your answer for the t-value is correct, and the t-value can be negative - it does not have to be positive, as you have incorrectly assumed. If the t-value is negative, it means that the mean of the second dataset is greater (i.e. your Lab 2 results are greater than your Lab 1 results), and you can still calculate the p-value.
You simply evaluate the Cumulative Distribution Function (CDF) of the t-distribution from $-\infty$ to $-1.5621$ (so you should not ignore the negative sign) for degree of freedom of $18$ (number of samples of Lab 1 plus number of samples of Lab 2 minus 2). You can use a table, or one of the many online calculators available on the web to carry out this operation. Let's call the result you obtain $q$. Assuming a two-tailed test, you multiply this result by 2 to obtain the p-value, so that the p-value is $2\times q$.
Supposing you had a positive t-value. You would would need to evaluate the t-distribution CDF from the t-value to $+\infty$, for the appropriate value of the degree of freedom. Calling this result $q$, the two-tailed p-value value would be $2\times(1-q)$.