Supposed that you are interested in estimating country-level maternal mortality rate (mmr_t) based just on the gross domestic product per capita (gdppc_t) and literacy rate (lr_t) and you find that countries that have unusually high (for the given levels of gdppc and lr) mmr in one period also have unusually high mmr in the next period. Which of the following assumption for time series analysis does not hold? A)No perfect collinearity. B)Zero conditional mean. C)Homoskedasticity. D)No serial correlation.

Question

Quizplus · Accepted Answer

The assumption of no serial correlation does not hold in this case, as the high mmr in one period is correlated with high mmr in the next period. This suggests the presence of a trend or systematic pattern in the data over time, which violates the assumption of no serial correlation. The other assumptions (no perfect collinearity, zero conditional mean, homoskedasticity) may still hold in this scenario.

Supposed That You Are Interested in Estimating Country-Level Maternal Mortality