We use Monte Carlo simulations and real data to assess the performance of alternative methods that deal with measurement error in investment equations. Our experiments show that individual-fixed effects, error heteroscedasticity, and data skewness severely affect the performance and reliability...