Data Analysis & Statistical Methods Flashcards

How many runs does a full factorial design with 3 factors at 2 levels each require, without replicates?

Answer: 8

A two-level full factorial needs 2^k runs, and 2^3 = 8.

Which design is most suitable for finding the optimal levels of acid and sweetener in a beverage flavor, including curvature?

Answer: Response surface methodology, such as a central composite design

RSM designs such as the central composite design fit quadratic models that locate an optimum.

On a Shewhart control chart for a flavor QC attribute, where are the control limits usually set?

Answer: Plus or minus 3 standard deviations from the mean

Shewhart charts use limits at 3 sigma, which contain about 99.7% of common-cause variation.

A flavor blend has three components whose proportions must sum to 100%. Which experimental design is appropriate?

Answer: Simplex mixture design

In a mixture design the proportions are constrained to a fixed total, so they are modeled on a simplex.

What is the purpose of a Williams Latin square serving design in sensory testing?

Answer: To balance presentation order and first-order carryover effects

In a Williams design, each sample precedes every other sample equally often, which balances carryover.

A test gives p = 0.03 at a significance level of alpha = 0.05. What is the correct conclusion?

Answer: Reject the null hypothesis

When p is below alpha, the null hypothesis is rejected at that significance level.

The same 12 panelists rate the bitterness of both a control and a test sample. Which t-test is appropriate?

Answer: Paired (dependent) t-test

Ratings from the same subjects are dependent, so a paired t-test on the within-panelist differences is correct.