A company which sells biscuits claims “Contents 880 grams” on the package. A sample of 28 packages yields an average of 800 grams. From past experience, the population standard deviation has been 50 grams. Using a .05 significance level, what conclusions would be drawn concerning the standard which the company is trying to achieve?
Null hypothesis: "u=880"
Alternative hypothesis: "u\\not =880"
Since the sample size is small(<30), it is appropriate to use t-statistics
Also, the alternative hypothesis is "u\\not =800", so two-sided test is required
Test statistic: "T(27)={\\frac {(800-880)*\\sqrt{28}} {50}}=-8.46"
Critical value can be found in tables for t-statistic, it is equal to -2,052
Since value of test statistic is (<0) and smaller than critical value, it means that null hypothesis should be rejected.
There is a significant evidence that company claim is not true, the real mean weight is smaller than company claims.
Comments
Leave a comment