|
|
 |
| |
|
|
|
Measures of Sample Skewness and Kurtosis for AR (1) Model with Missing Data with Applications in Economic Data |
|
|
|
PP: 1045-1056 |
|
|
|
doi:10.18576/jsap/130316
|
|
|
|
Author(s) |
|
|
|
Mahmoud M. Abdelwahab,
|
|
|
|
Abstract |
|
|
| Many statistical functions require that a distribution be normal or nearly normal. In time series models, data is assumed to follow a normal distribution. Two numerical measures of shape skewness and kurtosis can be used to test for normality. Autoregressive models are one of the models used to estimate time series data. The missing data in time series models affect in terms of their following a normal distribution. In this paper, derive the moments, skewness and kurtosis of AR (1) Model with Missing data without constant term and used the parameter of the model by using ordinary least squares (OLS), Yule Walker (YW) and weighted least squares (WLS). Moreover, Monte Carlo simulation at various sample sizes and different proportions of missing data for comparative study skewness and kurtosis between ordinary least squares (OLS), Yule-Walker (YW) and weighted least squares (WLS). In addition, time series real data with missing data was measures of kurtosis could be used to compare between these methods.
|
|
|
|
|
|
|
|
