I remember that I made such as this topic. I think I examined 30 days before and after announcement of dividend payout. You use event study method or Mann Whitney U method this method examined significant difference between two variables or events. 

I am agree with Prof. Süleyman Serdar Karaca,  and we  can change the days as per financial market  condition of the country. 

You can use questionnaire method.

If the question is only about measurement - you should probably use some factor and/or cluster analysis (I am fan of classification trees lately), where you should first find the significance and level of contribution of the earnings announcement among the different factors.

If you are looking into predicting the effect - I have seen many presentations where a word processing algorithm (typically some low-level AI) is used to process mass data sets (millions of words from public forums, twitter, news sites, etc.) to predict the price change.

The Event study method first applied by FFJR is traditionally used in these types of studies. I have personally used it to look at effects of stock splits on market capitalization. I would not advice use of 60 days (30 before, 30 after). You need some reasonable time.....like a year before and after. please read Econometrics of Event Studies by S.P. KOTHARI and JEROLD B. WARNER, in the book Handbook of Corporate Finance (Vol. 1) edited by B. Espen Eckbo (Publisher: North Holland, 2007) Quite an easy one to run, though. The best.

Event study following the methodology FFJR. With regard to the time horizon, depends on the degree of informational efficiency of the stock market that you want to investigate.

Yup, I agree with Prof. John. Please go through the Handbook of Corporate Finance (Vol.1). In my studies on M&A and stock market performance, I used both CAR and BHAR for the short-run and long-run performance.

You can use Cumulative Abnormal Ruturn (CAR) method to evaluate earnings announcement effect on stock price.

n stocks, the sum of all the differences between the expected returns and the actual returns up to a given point in time. Since the expected return is computed by an asset pricing model, the cumulative abnormal return may be used to determine how accurate the model is. More often, it is used to investigate the affect extraneous events have on stock prices.

Sum of the differences between the expected return on a stock (systematic risk multiplied by the realized market return) and the actual return often used to evaluate the impact of news on a stock price.

http://schwert.ssb.rochester.edu/f423/Data_mukim0803.xls

You can use efficient market hypothesis, event study method. To determine a window period for analyzing the stock reflection available information (earning announcement), You had better consider the level of market efficacy and the degree of information efficiency of the stock market. Some researchers choose 10 days lag and 10 days lead window period for their event studies depend on the market situation and information efficiency existing in the respective countries. And  I agree all the above discussions.    

You could use event study method, but yuo have to identify a precise announcement date (critical event).  For further information about the method I show you a brief description of this method, comes from my paper:

"Following Dodd and Warner (1983) and Brown and Warner (1985) we apply event-study methodology; it is a commonly used methodology to measure the stock market reaction to the announcement of a particular event (in this case, the critic event is the closing date of M&As).

The strong theoretical foundation for the event study methodology is the Efficient Market Hypothesis (EMH) (Fama et al., 1969; Fama, 1970) that define a market efficient if “prices fully reflect all available information”. The return-generating model we specify is the Market Model (Sharpe, 1963) which relates the return of a security to the return of market index as shown below:

1)Rit=αi+βi R mt+εit,

where Rit is firm’s i stock return at time t, Rmt is the market return, and εit is an i.i.d. normally distributed error term. Values for the model’s parameters αi and βi are estimated using the Ordinary Least Squares (OLS) method. In particular, αi measures the mean return over the period not explained by the market, while the βi measures the security’s i sensitivity to the market.

The following step is the estimation of Abnormal Return (AR) (2) defined as the difference between the Actual Return on a stock i and the Expected Return on the stock i:

2)ARit=Rit-(αi+βi Rm t),

where Rit and βi Rm t are respectively the bidder’s i daily return at time t and the expected return as computed in the equation (1).

Next we compute the Cumulative Abnormal Return (CAR) between any two dates T1 and T2 as:

CARi (T1,T2)=∑_(t=T1)^T2▒ARit,

and the Average Cumulative Abnormal Return (ACAR) as follows:

ACARi (T1,T2)=1/N ∑_(t=T1)^T2▒ARit ".

We could also use modified market model that doesn't require regression analysis.

You can use event study analysis and CAR method to evaluate earnings announcement effect on stock prices like dividend announcement effect to stock price reaction. You must use multiple event windows, depending on financial market efficiency (you can see also our article in Researchgate: Do Dividend Announcements Affect The Stock Prices in The Greek Stock Market? Vazakidis Athanasios, Athianos Stergios

International Journal of Economic Sciences and Applied Research. 01/2010)

Greeting

I am sending you work on the same subject in the appendix. Work is in Croatian, so use Google Translate for its transformation into English.

i think this study has been multiple times

i think this study has been done multiple times

Dear Ahmad Mohsen. You may consider the value relevance methodology by Ohlson (1995) I.e. Market price = a + b1 Book value persuade + b2 Earnings pershare.

You may separate the model into Market price = a + b1 Book value pershare and

Market price = a + b1 Earnings pershare.

HI,

This paper could give you some relevant ideas:

Earnings announcements and option returns

Journal of Empirical finance, January 2017, Pages 220-235

Sung Gon Chung, Henock Louis

Best regards