preview

Differences In A Meta-Analysis Study : An Analysis Of Varia

Decent Essays

Following Shadish, Cook and Campbell (2002), there are two standard methods to calculate effect size measures in a meta-analysis study, depending on the scale of the outcome variables. The standardized mean difference statistic (d) is used to computer the effect size for continuous outcome variables, whereas it employs the odds ratio for categorical outcome variables. The standardized mean difference statistic is calculated as: d_i=(X ̅_i^t-X ̅_i^c)/s_i (1) where i is study ith, X_i^t denotes the mean of the treatment group, X_i^c represents the mean of the control group, and s_i is the pool standard deviation of the two groups. 5.2 Instrumental Variable (IV) There are a number of methodological issues complicated the …show more content…

The effect of the disaster on human capital is estimated through divergence in educational attainment between the two groups after the tsunami. The general idea of SCM is explained below. Suppose that we observe 26 provinces in Indonesia for the period t=1980,…,2004,…,2016. Let i=1 be the province of Aceh, and i=2,…,26 be the other provinces that serve as the potential controls for Aceh. Here, we let T_0=2004 be the year when tsunami struck Aceh. We denote Y_it^I as educational attainment in the presence of the tsunami, while Y_it^N is educational attainment if the tsunami had not occurred. It is generally acceptable to assume that the disaster does not have any effects on the path of educational attainment prior to its occurrence at time T_0. Hence, Y_it^I=Y_it^N for t∈[0,…,T_0-1]. The human capital effect of the tsunami for province i at time t is written as α_it=Y_it^I-Y_it^N (3) Suppose D_it is a binary variable that takes a value of one if province i is exposed to the tsunami at time t and zero otherwise. The post-tsunami outcome for province i at time t can be represented as: Y_it=Y_it^N+α_it D_it (4) In this case, the Indonesian province of Aceh is the only province that severely hit by the tsunami after T_0. Therefore, D_it={█(1 if i=1 and t>T_0 @0 otherwise )┤ The goal is to

Get Access