Three topics of statistical methods for non-precise information are investigated in fuzzy environment as follows: 1. Fuzzy confidence intervals, 2. Testing statistical hypotheses, and 3. Fuzzy regression. For the first topic, we introduce an approach to obtain fuzzy confidence interval for fuzzy parameter based on fuzzy random variables. For the second topic, we investigate two approaches to testing fuzzy hypotheses as follows: · A fuzzy confidence interval based approach, which obtains a fuzzy test function using the relationship between confidence intervals and testing hypotheses, and · A fuzzy test statistic and fuzzy critical value based approach, which compare these values using a criterion to obtain a fuzzy test function. Finally, the problem of fuzzy regression for crisp input-fuzzy output, and also for fuzzy input-fuzzy output is studied. In this regard, three methods are proposed. MSC 2010: 62A86, 62C86, 62F03, 62F10, 62F15, 62G07, 03E72 Key words: Degree of acceptance (DA), Degree of rejection (DR), Fuzzy confidence interval, Fuzzy critical value, Fuzzy hypothesis, Fuzzy test statistic, Fuzzy robust regression, Testing hypothesis, Variable read fuzzy regression model, Most powerful fuzzy test, Uniformly most powerful fuzzy test, Least- absolutes fuzzy regression, Least-squares fuzzy regression.
