Question

I would love if you show the procedure fore :  comparing the difference in mean for two different chips brands which I consume. I will see if the mean pieces of chips for an American Brand is less, higher or equal to that of an Canadian.

Explanatory variable: The number of pieces of ships

Response variable: The outcome of the study ( which has higher, less or are they equal)

what exactly I should do avoid bias and what is the proper way to gather data. It is just stimulation and doesn’t have to be actual data but brings a possible outcome with predicted values by you.

Answer 1

To collect data avoiding bias we have to keep attention on following facts.

• We have ot collect sufficient enough representative of both population.
• During collecting data, we have to collect records corresponding to products over longer length of time so that there is no bias due to sampling.

We have to get a good representative of whole population. To do so we have to proceed as follows.

• We have to access record over different lots of chips. This is to be done over time. So, we have to collect data regarding products over sufficiently longer time (e.g. for one month or two).
• From different factories of those brands, there may be difference in production. So, we have to include all production houses of that brand.

Thus over reasonably longer (but not too much, as it would be costly) time period we have to gather sample packets of chips from all factories of certain brand and from all lots that are produced and about to deliver in different places.

After collection of data we have to perform hypothesis testing as follows.

Suppose, random variables X and Y denote number of chips in packets of American brand and Canadian brand respectively.

Here, brands of two different countries are used to collect data to compare between number of chips. Further we do not know population standard deviation (or variance). So, we have to perform two sample t-test.

We have to test for null hypothesis

against our required alternative hypothesis or or