|
|
|
||||
|
|
|||||
Figure 3.1, page 87.
(a)
proc print data = pain; where paintol = .; run; data pain; set pain; where id not in (4, 18, 20, 33, 42, 44); run; proc sort data = pain; by id cs; run; proc means data = pain; where trial ne 4; by id cs; var l2paintol; output out = pain2; run; data pain2; set pain2; if _STAT_ = "MEAN"; rename _STAT_ = l2paintol; keep l2paintol id cs; run; proc univariate data = pain2 noprint; where cs = "attender"; histogram l2paintol / midpoints=3.167 to 7.833 by .333 vscale = count; run;
(b)
proc univariate data = pain2 noprint; where cs = "distracter"; histogram l2paintol / midpoints=3.167 to 7.833 by .333 vscale = count; run;
Table 3.1, page 88.
proc ttest data = pain2; class cs; var l2paintol; run;
The TTEST Procedure
Statistics
Lower CL Upper CL Lower CL Upper CL
Variable cs N Mean Mean Mean Std Dev Std Dev Std Dev Std Err
l2paintol attender 29 4.1384 4.473 4.8075 0.6979 0.8795 1.1895 0.1633
l2paintol distracter 29 4.6601 5.0762 5.4923 0.8681 1.0939 1.4794 0.2031
l2paintol Diff (1-2) -1.125 -0.603 -0.081 0.8379 0.9925 1.2175 0.2606
T-Tests
Variable Method Variances DF t Value Pr > |t|
l2paintol Pooled Equal 56 -2.31 0.0243
l2paintol Satterthwaite Unequal 53.5 -2.31 0.0245
Equality of Variances
Variable Method Num DF Den DF F Value Pr > F
l2paintol Folded F 28 28 1.55 0.2544
Table 3.2, page 89.
data transform; input l2seconds; datalines; .6 .88 1.09 4.47 5.08 3 4 5 6 7 8 ; run; data transform; set transform; seconds = 2 ** l2seconds; run; proc print data = transform; run;
Obs l2seconds seconds 1 0.60 1.516 2 0.88 1.840 3 1.09 2.129 4 4.47 22.162 5 5.08 33.825 6 3.00 8.000 7 4.00 16.000 8 5.00 32.000 9 6.00 64.000 10 7.00 128.000 11 8.00 256.000
Figure 3.2, page 90.
(a) Follows formula from page 89
proc means data = pain; var l2paintol trial; run; data pain; set pain; ystar = l2paintol - 4.7730379; tstar = (trial - 2.5); tstar2 = (tstar) * (tstar); ytstar = ystar * tstar; run; proc sort data = pain; by id treatment cs; run; proc means data = pain sum; by id treatment cs; var ytstar tstar2; output out = pain3; run; data pain3; set pain3; where _STAT_ = "MEAN"; drop _FREQ_ _STAT_ _TYPE_; ytstar = 4 * ytstar; tstar2 = 4 * tstar2; slope = ytstar / tstar2; run; proc univariate data = pain3 noprint; where cs = "attender"; histogram slope / midpoints=-.9375 to 0.9375 by .125 vscale = count RTINCLUDE; run;
(b)
proc univariate data = pain3 noprint; where cs = "distracter"; histogram slope / midpoints=-.9375 to 0.9375 by .125 vscale = count RTINCLUDE; run;
Table 3.3, page 90.
proc ttest data = pain3; class cs; var slope; run;
The TTEST Procedure
Statistics
Lower CL Upper CL Lower CL Upper CL
Variable cs N Mean Mean Mean Std Dev Std Dev Std Dev Std Err
slope attender 29 -0.109 -0.017 0.0739 0.1904 0.2399 0.3245 0.0446
slope distracter 29 -0.187 -0.046 0.0959 0.2954 0.3722 0.5034 0.0691
slope Diff (1-2) -0.136 0.0283 0.193 0.2644 0.3131 0.3841 0.0822
T-Tests
Variable Method Variances DF t Value Pr > |t|
slope Pooled Equal 56 0.34 0.7324
slope Satterthwaite Unequal 47.8 0.34 0.7326
Equality of Variances
Variable Method Num DF Den DF F Value Pr > F
slope Folded F 28 28 2.41 0.0234
Figure 3.3, page 91.
data pain3; set pain3;
if cs = "attender" then do;
if treatment = "attend" then cst = "AA";
else if treatment = "distract" then cst = "AD";
else cst = "AN";
end;
if cs = "distracter" then do;
if treatment = "attend" then cst = "DA";
else if treatment = "distract" then cst = "DD";
else cst = "DN";
end;
run;
proc sort data = pain3;
by cst;
run;
proc boxplot data = pain3;
plot slope * cst;
run;
Table 3.4, page 92.
data t2; set pain;
where trial = 2;
keep id cs treatment l2paintol;
run;
data t4; set pain;
where trial = 4;
keep id cs treatment l2paintol;
run;
data t2; set t2;
rename l2paintol = l2pt2;
run;
data t4; set t4;
rename l2paintol = l2pt4;
run;
proc sort data = t2;
by id;
run;
proc sort data = t4;
by id;
run;
data pain24; merge t2 t4;
by id;
if cs = "attender" then do;
if treatment = "attend" then cst = "AA";
else if treatment = "distract" then cst = "AD";
else cst = "AN";
end;
if cs = "distracter" then do;
if treatment = "attend" then cst = "DA";
else if treatment = "distract" then cst = "DD";
else cst = "DN";
end;
run;
proc sort data = pain24;
by cst;
run;
proc ttest data = pain24;
paired l2pt4 * l2pt2;
by cst;
run;
[...output omitted...]
Figure 3.4, page 93.
data pain24; set pain24; diff = l2pt4 - l2pt2; run; proc sort data = pain24; by cst; run; proc boxplot data = pain24; plot diff * cst; run;
Table 3.5, page 93.
data AAAN DDDN; set pain24;
if cst in ("AA" "AN") then output AAAN;
if cst in ("DD" "DN") then output DDDN;
run;
proc ttest data = AAAN;
class CST;
var diff;
run;
The TTEST Procedure
Statistics
Lower CL Upper CL Lower CL Upper CL
Variable cst N Mean Mean Mean Std Dev Std Dev Std Dev Std Err
diff AA 10 -0.445 0.1819 0.8085 0.6025 0.8759 1.599 0.277
diff AN 9 -0.393 0.0284 0.4501 0.3706 0.5486 1.051 0.1829
diff Diff (1-2) -0.564 0.1536 0.871 0.5554 0.7401 1.1096 0.3401
T-Tests
Variable Method Variances DF t Value Pr > |t|
diff Pooled Equal 17 0.45 0.6573
diff Satterthwaite Unequal 15.3 0.46 0.6501
Equality of Variances
Variable Method Num DF Den DF F Value Pr > F
diff Folded F 9 8 2.55 0.2022
proc ttest data = DDDN; class CST; var diff; run;
The TTEST Procedure
Statistics
Lower CL Upper CL Lower CL Upper CL
Variable cst N Mean Mean Mean Std Dev Std Dev Std Dev Std Err
diff DD 11 0.0682 0.7153 1.3625 0.6731 0.9633 1.6906 0.2905
diff DN 8 -1.8 -0.679 0.4427 0.8868 1.3413 2.7299 0.4742
diff Diff (1-2) 0.282 1.394 2.5061 0.8512 1.1343 1.7005 0.5271
T-Tests
Variable Method Variances DF t Value Pr > |t|
diff Pooled Equal 17 2.64 0.0170
diff Satterthwaite Unequal 12 2.51 0.0275
Equality of Variances
Variable Method Num DF Den DF F Value Pr > F
diff Folded F 7 10 1.94 0.3303
[http://www.ats.ucla.edu/stat/sas/footer.htm]