@nils_weidmann @the_linde ¡Smart approach for measuring inequality! However, #validity concerns arise. Combined indicators in this #analysis can throw inconsistencies. #NTL can depend on factors like overall lack of #infrastructure in communities, and not so much on individual inaccessibility #ieMethods
@Marc_S_Jacob @DominikSchraff ¡Interesting approach! Locating divergent understatings through a #conjointanalysis helps trace inconsistencies in #political affiliations. By evaluating "democratic" standings, one can uncover attributes that influence populations' decisions and accepted trade-offs #ieMethods
@AndreasShrugged@UNUWIDER#Panel structure #analysis is most useful in this case, as as the effects and differences between age and cohort cannot be observed in a single cross-sectional. Could the results also apply to other developing countries, in Asia or South America, for example? #ieMethods
@AndreasShrugged@UNUWIDER#Confounding variables that may alter the results –religious beliefs, social norms, or violence– are particularly important. Also, labor participation does not indicate growth in income, or professional permanence in any sector. This would be interesting to evaluate #ieMethods
@IEGPAMIRSS Along the same lines, #confounding#variables may exist that influence students' #food decisions, not directly related to food availability. Participants in an #experimental study might be influenced by #extraneous factors that are outside of the researchers' control. #ieMethods
@IEGPAMIRSS#Experimental trials have the advantage of being combinable with other studies, as we see in this case. However, under such tight controls, unrealistic situations can be created, as the data produced can be corrupted or inaccurate, whilst still appearing authentic. #ieMethods
In the current global context, the belief/rejection of #fakenews can determine the support/future of a country. It would be interesting to check for #externalvalidity, and measure if ongoing war conflicts have any direct effect on the results #ieMethods https://t.co/Fl9jE4xncZ
@simonweschle @Sarah_Brierley1 @chrishanretty Interesting application of #DiD design. Useful when randomization is unfeasible or complicated, but causal inference may present an important limitation in the results, as general assumptions may not apply to individual cases #ieMethods
@DominikSchraff Great example of #RDD. Estimation of #bandwidth is tricky and a critical #design choice. Important to make sure treated and untreated units are similar on all other variables, except for the running (vote margin). This makes an #empirical assessment possible. #ieMethods
@monicambravo@BarbaraBiasi@rema_nadeem Using the potential of structural modeling and RCTs to answer economic/development questions. #Dynamism: combining a structured, cost-saving approach, with bias minimization and statistical reliability to measure #policy#impact in a holistic way 🙌👏 #ieMethods
Though often seen as rivals, #RCT and #structural#estimation can complement each other in a variety of ways; specially when measuring #policy impacts in #development fields. "The Best of Both Worlds: Combining RCTs with Structural Modeling" #ieMethods https://t.co/9ljLSh4cwY
School enrollment in Mumbai and Delhi has risen, but the government shows inabilities to retain and adequately #educate children; why? #Field#studies have shown that the reason has little to do with their families' economic circumstances #ieMethods
https://t.co/hHCJXSQ7yR
@abigailfrings Interesting topic. Especially because it is not based exclusively on U.S. #aid allocation, but integrates the 21 member countries of the @OECD Development Assistance Committee. This gives the study greater #external#validity, making it applicable in multiple scenarios #ieMethods