STAT-2126EL Lecture 6: Introduction to the t Statistic

• School: Laurentian
• Department: Statistics / Statistique
• Course: STAT-2126EL
• Professor: Emond Michael

Document Summary

Now we move on to another test statistic called the t score. Since we can"t get the population standard deviation we go for the next best thing: the sample variance. We can now use this to calculate an estimated sample error (standard error) by plugging it into the old standard error formula. It is an estimate of the standard difference (expected by chance) between our sample mean and our hypothesised population mean. We can now use out derived values of standard error and \mew and substitute them into our z score formula to have our new t score formula. We had to hypothesize a population mean without any actual score. We had to estimate the standard error using our sample variance, because we had not \sigma (standard deviation) from the actual population. The t statistic is used to test hypothesis about \mu when the value for \sigma is not known and you can determine a reasonable value for \mu.