About Newey-west (HAC) results

[YANGXIAO]

I find a standard eviews result from an article, which as follows:

Dependent Variable: RATES
Method: Least Squares
Date: 12/03/12 Time: 14:09
Sample: 1961Q1 1986Q3
Included observations: 103
HAC standard errors & covariance (Prewhitening with lags = 1,
Quadratic-Spectral kernel, Andrews bandwidth = 1.9610)
Variable Coefficient Std. Error t-Statistic Prob.
C 1.375142 0.599818 2.292600 0.0239
R-squared 0.000000 Mean dependent var 1.375142
Adjusted R-squared 0.000000 S.D. dependent var 3.451231
S.E. of regression 3.451231 Akaike info criterion 5.325001
Sum squared resid 1214.922 Schwarz criterion 5.350580
Log likelihood -273.2375 Hannan-Quinn criter. 5.335361
Durbin-Watson stat 0.745429

My eviews result as follows:

Dependent Variable: PEMA_W
Method: Least Squares
Date: 05/12/14 Time: 22:59
Sample: 5/13/2013 5/05/2014
Included observations: 247
HAC standard errors & covariance (Bartlett kernel, User bandwidth =
8.0000)

Variable Coefficient Std. Error t-Statistic Prob.

C 772.7935 232.9400 3.317565 0.0010
SEMA_W 157.8913 38.04165 4.150486 0.0000

R-squared 0.143515 Mean dependent var 1749.567
Adjusted R-squared 0.140019 S.D. dependent var 86.53615
S.E. of regression 80.24943 Akaike info criterion 11.61622
Sum squared resid 1577793. Schwarz criterion 11.64464
Log likelihood -1432.603 Hannan-Quinn criter. 11.62766
F-statistic 41.05288 Durbin-Watson stat 0.049852
Prob(F-statistic) 0.000000 Wald F-statistic 17.22653
Prob(Wald F-statistic) 0.000046

I have two questions:
(i) why are the two eviews results different by using the same method (focusing on the parts of the underlined and bold)? How could I get the standard result as the article result shown by running eviews?
(ii) I use this eviews method in order to remedy auto-correlation, but my eviews result does not seem to be useful to remedy auto-correlation. How could I reach my goal by using eviews?

Thank you very much!
Hope your response asap!

[EViews Glenn]

HAC computation doesn't remove autocorrelation. It corrects standard errors for heteroskedasticity and autocorrelation.

You need to set the HAC settings to match. Click on the HAC options button.

[YANGXIAO]

I click the estimation first, and click the option, then choose the coefficient covariance matrix to HAC, then do the estimate, but from the result as i show you yesterday, i cannot get the same result as the 1st result from the article, specifically speaking, the black sentences are different in the two results, i wanna know how to get the first one.

[EViews Glenn]

Once you choose HAC as your covariance method, there will be a HAC options button that you have to press which brings up another dialog. You'll have to set the options to match the results you wish to match. You are using a different kernel and bandwidth, and the results you are trying to match also do prewhitening.