Volume : 6, Issue : 5, May - 2017

Jackknife Method to Estimate the Confidence Intervals about an R-squared statistic Using SAS Software

Hasan Talib Hendi

Abstract :

<p>&nbsp;<span style="font-size: 12pt; line-height: 150%; font-family: &quot;Times New Roman&quot;, serif;">R</span><span style="font-size: 12pt; line-height: 150%; font-family: &quot;Times New Roman&quot;, serif;">e-sampling methods cover a range of ideas. All of them involve taking a sample and using it as our basis for drawing new samples, hence the name &lsquo;Resampling&rsquo;. We sample from our sample, in one way or another. There are several ways in which we can &lsquo;resample&rsquo;. We can draw &lsquo;randomly&rsquo; from our sample, or we can draw designed subsets from our sample. Also, we can pull the data but exchange the labels on the data points. For example in bootstrapping, we want to approximate the entire sampling distribution of some estimator. We can do that by resampling from our original sample.</span></p> <p class="MsoNormal" style="line-height:150%"><span style="font-size:12.0pt;&#10;line-height:150%;font-family:&quot;Times New Roman&quot;,serif">In this dissertation we focus on generating samples by resampling techniques by jack-knifing, perform regression analysis for generated samples and then estimate the confidence intervals about an R squared statistic by using SAS software^[1].<o:p></o:p></span></p> <p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height:&#10;150%"><b><span style="font-size:12.0pt;&#10;line-height:150%;font-family:&quot;Times New Roman&quot;,serif;mso-fareast-font-family:&#10;Calii">&nbsp;</span></b></p>