Paired samples or independent sample hypothesis test for two time periods I want to know if employees in an organization are surveyed in 2013 and again in 2014 if the samples from the two time periods are considered related and dependent or independent if they are asked the same likert-type questions for purposes of a hypothesis test for change on individual items. They would not be matched pairs. For small samples what would be the appropriate statistical test and should a finite population factor be considered for confidence intervals?
 A: A standard chi-squared test or a Fisher's exact test (better for small samples) will test the independence of Likert scores in 2013 vs. 2014. For 2013-2014 comparisons on particular items, simple binomial tests with Tukey or Bonferroni significance level corrections should be fine. 
If the same employees are completing the survey in both 2013 and 2014, but you cannot identify employees by name or ID number, then the power to detect a general change in Likert scores will be reduced. 
However, if it is possible to match 2013 and 2014 responses by employee ID, then use the Stuart-Maxwell test as a replacement for the chi-squared test for matched pairs where there are more than two possible categories for outcome data. 
If there are other predictor variables aside from year (such as age, gender, job classification) in your data, I suggest a multinomial or ordinal logistic regression.
You don't need to use a finite population correction for hypothesis tests, only for univariate descriptive statistics. Hypothesis tests assume you are always sampling from a near infinite "super-population" for which you are making inferences.
