of Clustered Data Thai investors are classified into three clusters using investment proportions (Proportions of their money allocated to each investment type) Cash Holder 444 20% Old-Fashioned 1,573

allocated to each investment type e.g. cash deposit, stock market, real estate, gold, TFEX, and T-bond) as clustering variables. Three investor clusters are uncovered, namely, cash holders, old-fashioned

emissions per p-km and t-km than other modes of travel. Table 4 shows various relevant emissions estimates from the literature for rail and road. From these figures it can reasonably be inferred that direct

; Madueno et al., 2016; Nicolosi et al., 2014). A latent variable is not directly observed but rather inferred from other proxies that can be measured. Given several sub-varieties of latent variable analysis