I was hoping I could get some beneficial input as to what problem I should address for my statistical quality control project. As of now we have covered inferences on samples and statistical quality/process control charts.

From the class instructions:
The student will identify an application area of interest, clearly specify the study question, propose how to study the question empirically, develop a data collection methodology, collect the raw data herself/himself, select and justify the use of a specific model or models as appropriate, implement and evaluate the model as appropriate, critically analyze the data including appropriate and well-designed plots to determine what can be legitimately concluded, and effectively report the results in oral and written report format.

Application areas are of three general types. First is the application of a statistical process control model to some process (material from the first part of the course). Second is the conduct of an experiment (material more from last part of course) to identify and quantify potential cause-and-effect relationships between a performance measure and one or more independent variables. The third is the conduct of a descriptive empirical study to gain deeper understanding of a current condition to propose what should be done next.

If you do an A vs B experiment, you cannot be a subject in the experiment.

You are not allowed to do two types of projects: 1--a comparison of grocery store self-checkout vs a clerk-checkout, and 2a comparison of alternative commuting routes to and from school.



I was thinking about something like testing torque sensors and seeing how many in a certain batch fall within capacity specification limits and what is the true average (I work at a torque sensor manufacturing company). Also I seen on the internet an idea where we test m&m's and see if the weight depends on color, but not sure about that one. Also I saw one about testing which chicken noodle soup brand has more chicken.

If you guys have any other good ideas that would be more applicable to my project I would be grateful.