I would like to know the latest tools used to measure stock market efficiency because a wide variety of statistical tests have been used in the literature to examine the validity of weak form of efficient market hypothesis.
Indeed, there is a wide variety of statistical tools for examining the weak form of EMH. Finally, you can never say a market is efficient, even in its weak form, but there are no reason to reject EMH...
I think a very good survey is made by Lim & Brooks - THE EVOLUTION OF STOCK MARKET EFFICIENCY OVER TIME: A SURVEY OF THE EMPIRICAL LITERATURE, Jounal of Economic Surveys, 2011, Vol. 25, No. 1, pp. 69–108.
Also, together with my colleague, Elena Tilica, I count for a list of some tests (which is not exhaustive) here:
In addition to the standard academic tools, you might wish to consider how the U.S. courts have approached this subject. There are billions of dollars at stake (and sometimes criminal sentences) when the courts are involved. An easy entrance to this body of law and analysis is to use the search terms "Camer factors" and 'fraud on the market'. And thanks, Victor, for the citations.
According to Fama (1970), an asset market is efficient if the asset price fully reflects all available information. There are at least three models for testing stock market: the Fair Game property (a stochastic process Xt with the condition in information set It), the Submartingale model (Fair Game model with a small adjustment in expected return; the expected return is considered to be positive instead of zero and the small adjustments mean that prices are expected to increase slightly over time) and the Random Walk model (the random walk testing is more accurate and most convenient).
The most common methods used to test stock market weak-form efficiency are the runs test (Bradley, 1968, Wallis and Roberts, 1956), the autocorrelation test (Durbin and Watson, 1951, Ljung and Box, 1978), the variance ratio test (Lo & Mackinlay, 1988), etc.