Hypothesis testing procedures compare a conjecture we have about a population to the information contained in a sample of data
Given an economic and statistical model, hypotheses are formed about economic behavior.
These hypotheses are then represented as statements about model parameters
Hypothesis tests use the information about a parameter that is contained in a sample of data, its least squares point estimate, and its standard error, to draw a conclusion about the hypothesis
Hypothesis tests are statistical tests that are best left to the classroom. The test is not scientific as it does not answer a scientific question. It is a test for data and tells if the data are tending in the direction of the tested effect. It tells you nothing about the effect. The science is in how much supports the effect and what is necessary to describe and control the effect.
Hypothesis testing is an act in statistics whereby an analyst tests an assumption regarding a population parameter. The methodology employed by the analyst depends on the nature of the data used and the reason for the analysis.
Once he has elaborated his project or research protocol on a certain topic, and preferably if a research problem has already been raised and he has advanced an assumption or conjecture (Hypothesis) as a response; The next step is to design the experiment, the strategies, methods and measurement and evaluation instruments, as well as the Operational or Working Hypothesis: Null Hypothesis (H0) and Alternate Hypothesis (H1).
With this theoretical, methodological and instrumental arsenal, research is oriented and carried out in the field of reality, results are obtained, analyzed, evaluated and interpreted. All this last, is the Hypothesis Test.
In the most general sense, hypothesis testing is a decision framework to understand some parameter that defines a population. The parameter is thought to have some baseline value, which is the null hypothesis. The alternative hypothesis is a claim directly contradicting the null. We then define some test statistic that points in the direction of what the parameter might be (e.g., if you're trying to estimate the true mean you might look at the sample mean), derive what the distribution of that statistic would look like if the null were true, and define ranges of values for the statistic that would be reasonable if the null hypothesis were really the true state of affairs. Gather some data, compute the statistic, and see whether the statistic is in that range or not. Decide accordingly.
Hypothesis testing is to prove or disprove one’s argument statistically.
When someone (Person/Organisation/Company) makes a claim, we have two options
Agree to the claim
DO NOT agree to the claim
If we agree to the claim - NO action is required (No test is required)
If we challenge the existing claim/status quo, we have show evidence to the contrary. Hence, we resort to statistical testing. We formulate a hypothesis which defines the objective of our test.
NULL Hypothesis - Status Quo/Claim made
Alternative Hypothesis - What we want to prove/Objective of statistical test
Hypothesis testing is an act in statistics whereby an analyst tests an assumption regarding a population parameter. The methodology employed by the analyst depends on the nature of the data used and the reason for the analysis. Hypothesis testing is used to infer the result of a hypothesis performed on sample data from a larger population.
Excellent answers provided thus far. Another way of looking at null (H0) vs. alternate (Ha) hypothesis is like this: If, for example, you want to prove that the method (or process or material) you devloped is statistically significantly different (hopefully better) than the exisiting one, the onus is upon you via hypothesis testing (after data collection) that this is indeed the case. If you are unable to do that the null hypothesis stands. Usually the statistical signifiance level is set at a standard value, say of 95%. However, statistical significance is just one criterion. In other cases, one may want to assess the clinical significance of, say, a new medical treatment. If the new medical treatment results in saving a few more lives than the exisitng method, it may well be embraced as a clinically signifiacnt finding even if it did not muster the rigor of statistical signifiance.
Nothing is proved by hypothesis testing, particularly at the level of significance. Significance is necessary but is far from sufficient. Significance does nothing to quantify the effect measured, it indicates that the data are trending to support. This is necessary for deciding to pursue further testing, because if there is no tendency to support, there is no reason to pursue the effect.
There is no way you will achieve clinical significance without statistical significance.