i) Construct the QQ plot with the theoretical qq line; use the function qqline() in R. Based on that you can tell that the median standardized residual as well as the two standardized residuals that immediately follow, are

[ Select "smaller" or "larger"] than their expected values for a normal distribution.

data set:

COST	PAPER	MACHINE	OVERHEAD	LABOR
1102	550	218	112	325
1008	502	199	99	301
1227	616	249	126	376
1395	701	277	143	419
1710	838	363	191	682
1881	919	399	210	751
1924	939	411	216	813
1246	622	248	124	371
1255	626	259	127	383
1314	659	266	135	402
1557	740	334	181	546
1887	901	401	216	655
1204	610	238	117	351
1211	598	246	124	370
1287	646	259	127	387
1451	732	286	155	433
1828	891	389	208	878
1903	932	404	216	660
1997	964	430	233	694
1363	680	271	129	405
1421	723	286	146	426
1543	784	317	158	478
1774	841	376	199	601
1929	922	415	228	679
1317	647	260	126	378
1302	656	255	117	380
1388	704	281	142	429