Lab 10: Instructions and Questions for Excel 2003
Inferential Statistics and Hypothesis Testing
NAME______LAB TIME______LAB BLDG______
This lab will carry out hypothesis tests and will examine the conditions which result in rejection of the null hypothesis.
- Go to the course webpage and open the lab10data file.
- Suppose we are interested in a variable whose mean is supposed to be 100. Thedistribution of this variable is believed to be normal with astandard deviation= 0.8 mm based on previous surveys. We want to detect a shift in the mean in either direction.
- The null hypothesis should be:__Ho:______
- The alternate hypothesis should be:__Ha:______
- To examine the effect of sample size, we will use two sample sizes, n=30 and n = 6. The first sample of 30 observations is given in column B, and the first sample of 6 observations is given in Column K.
- Use A45 through A49 for labels as follows: A45: sample size, A46: sample mean, A47: test statistic, z, A48: 1-side P Value, A49:2-side P Value.
- In B45 enter =30 for sample size.
- In B46 use the AVERAGE function to calculate the sample mean for B10:B39.
- In B47 calculate the test statistic, z, using the formula: z = (sample mean – hypothesized mean) / (0.8/ sqrt of sample size). Enter = (B46-100) / (0.8/SQRT(B45))
- In B48 use the NORMSDISTfunction to calculate the 1-side P Value. Enter = NORMSDIST(B47).
- In B49 use the NORMSDIST function to calculate the 2-side P Value. Enter = 2*NORMSDIST(B47)
CALCULATE 7 NEW COLUMNS OF HYPOTHETICAL SAMPLES OF SIZE = 30:
- In Column C generate a new sample of n= 30 items using excel to deduct 0.1 from each data value in the column B. ie, in C10 use the formula: =B10-0.1. Copy this formula down to generate 30 new values.
- Repeat the above instructions to generate 6 new data samples in columns D through I except deduct 0.05 instead of .1 from the column on the left.
- Copy the formulas in B45:B49 across to column I.
- The sample means in line 46 should be 100, 99.9, 99.85, 99.80, 99.75, 99.70, 99.65, 99.60, and the corresponding test statistics and P Values should be descending also.
CALCULATE 8 NEW COLUMNS OF HYPOTHETICAL SAMPLES OF SIZE =6
- In Column K a sample of n=6 items is given. In Column L generate a new sample of size n=6 using excel to deduct 0.15 from each data value in column K.
- Repeat the above instructions to generate 7 new samples in columns M through S except deduct 0.10 instead of 0.15 from the column on the left.
- In K45 enter =6. Copy K45 across to columns L through S. Copy the formulas in I46:I49 over to columns K through S.
- The sample means from K46 through S46 should be 100, 99.85, 99.75, 99.65…..99.15.
- There should be two graphs between lines 51 and 102 showing the sample means and the associated P Values which you just generated. The second plot just expands the y axis of the first plot to aid in interpretation.
INTERPRETATION OF SAMPLES 1-8 WHEN SAMPLE SIZE = 30
- The first sample of n=30 in Column B has a mean = 100, exactly as expected in the null hypothesis. Samples 2 - 8 in Columns C through I have descending means. Note that the P Values are also descending from Column B through I.
- Remember that you can reject the null hypothesis only when the P Value is equal to or less than the significance level, α.
- Sample 1 in Column B supports which hypothesis? Null______Alternate______
- As the sample mean decreases, going from Column B toward Column I the null hypothesis becomes: Easier to believe______Harder to believe______
- If α = .10, and with a two-side Ha, rejection of the null hypothesis would begin with what sample mean? ______
- If α = .01, and with a two-side Ha, rejection of the null hypothesis would begin with what sample mean? ______
- If α = .10, and with a one-side Ha, rejection of the null hypothesis would begin with what sample mean? ______
- If α = .01, and with a one-side Ha, rejection of the null hypothesis would begin with what sample mean? ______
- In general, is it easier (more likely) or harder (less likely) to reject a 2-side hypothesis than it is to reject a one-side hypothesis? ______
COMPARISON OF SAMPLE SIZE N=30 WITH N=6
- Comparing sample size n=30 with n=6, for a given value of the sample mean, and for a given set of hypotheses, the P Value is always lower for which sample size? ______
- For detecting small shifts in the population mean, it would be better to use a large sample size or a small sample size? ______
COMPARISON OF α = .10 VS α = .01
- In general, which value of α would result in rejection of Ho more often, α = .10 or α = .01? ______
- Which value of α would be regarded as the most rigorous criteria (most difficult to achieve) for rejection of Ho, α = .10 or α = .01? ______
- The P Value represents the probability that the particular value of the sample mean (or a value more extreme) could occur by chance alone WHEN THE NULL HYPOTHESIS IS ACTUALLY TRUE. Knowing this, which value of α would result is more cases where Ho is rejected when Ho is actually true, ie: the mean is actually = 100, α = .10 or α = .01? ______
Print the spreadsheet including the plots of P Values, A1:S104 using landscape with a page break after line 43, and turn in with your answers to the questions.