Sanusi Shuaibu
Modified Jones model has been extensively applied and studied by scholars all over the world. There abundance of literature available.
1. First of all define your population and sample entities pre and post IFRS adoption period ( 2008 - 2012 and 2013-2017), it is prudent to avoid banking, finance, insurance and other entities which follows additional or special accounting treatments, as their accounting presentation is different for the rest of the entities.
2. Collect the annual report data in excel and clean and sort out the data. Further perform the beta value calculations in excel as well. Input the final mathematical values to the statistical software for analyzing.
3. It would be prudent to carry out residual testing on the data to further validate the model.
Please refer the links below
https://www.casrilanka.com/casl/images/2017/journal_of_applied_research_for_web.pdf
Article Board characteristics and earnings management in Sri Lanka
The following is a recent paper that used the modified Jones Model in this regard. You may want to check it out.
Effect of Mandatory Disclosure of IAS/IFRS on Earnings Management Among Listed Firms At The Uganda Securities Exchange
RO Etengu, TO Olweny, JO Oluoch - 2019 - ir.lirauni.ac.ug
You may use Collins at al. (2017) it's an extended version of performance adjusted model developed by Kothari (2005) which is also an advanced version of Modified Jones Model (1991).
Which of the modified Jones are you interested in? Dechow et al (1995) or Kaznisk (1996)? I suggest that you can use both. Because according to Peasenell (2000) they're the most reliable in detecting DA. Even Jones (1991) model is equally reliable when Dynamic panel such as GMM is use during estimation. You should note however, that the power to detect EM by the these models varies according to the countries laws and corporate governance codes.
I have been working on similar issue. In my case i have gone through all available jones type model in literature till 2017. Following that, it is documented that the extended modified jones model (Yoon et al., 2006) is more appropriate in the context of emerging economy or developing countries, like Hoang (2014) in case of Vietnamese firms; Alareeni and Aljuaidi (2014) in Palestine; Al-Rassas and Kamardin (2015) in Malaysia; Islam et al. (2011) and Islam (2014) in Bangladesh; (Yoon, Kim, & Woodruff, 2012) in Korea . You may consider that.