ࡱ>  n*bjbjvv 4B"+ \\\\\ppp8<,p<&&&`<b<b<b<b<b<b<,>Ҵ<\<y\\&&<yyy\&\&`<y`<yy659jsM7L<<0<7<AyA89y98\P9<<y<A ":   San Jose State University School of Social Work ScWk 242 Spring 2009 Lab Exercise #5: Logistic RegressionANSWER KEY You are comparing the effectiveness of group psychotherapy vs. individual psychotherapy for the treatment of depression, hypothesizing that group therapy will be as effective if not more so than individual psychotherapy. The sample consists of 50 adults diagnosed with moderate depression, chosen randomly from those referred to a large urban clinic. Clients were then randomized to an experimental group who received time-limited (15 weeks) group therapy each person assigned to one of five different therapy groups, or to a control group who received individual therapy for six months. The research design was an experimental treatment and control group, with pre- and posttest measurements of depression ((both groups measured posttest after six months). You are interested in comparing the effect of both treatments on whether or not clients mood and symptoms improved, controlling for ethnicity, having had a history of depression, whether or not there was parental history of depression, age, and current level of social support. [Note: this is a fictional data set for illustration purposes.] The variables included in the model are: Variable nameAs measured/coded byTreatment_groupTreatment group assignment: 1 = group psychotherapy 0 = individual psychotherapyDep_hxParticipants history of depression? 1 = Yes, 0 = NoAgeParticipants age in yearsParentParental history of depression? 1 = Yes, 0 = NoAsian_PIDummy variable for Asian or Pacific Islander, 1 = Yes, 0 = No (White is reference category)Afr_amerDummy variable for African American, 1 = Yes, 0 = No (White is reference category)LatinoDummy variable for Latino, 1 = Yes, 0 = No (White is reference category)High_supportDummy variable for self reported high level of support (compared to medium or no support). No support is the reference category. 1 = High social support 0 = Not high social supportMed_supportDummy variable for self reported some level of support (compared to high or no support). No support is the reference category 1 = Medium level of support 0 = Not medium level of supportImprovedMood and symptoms improved vs. not improved (summarized from a depression scale with cutoff score) 1 = Improved 0 = Not improvedTable 1 compares the two groups on all variables. [Notethis would normally go in the Sampling section of the Methods.] Table 1. Group comparisons on all variables. Group therapy (n and % of group total)Individual therapy (n and % of group total)Client history of depression12 (48%)12 (48%)Parental history of depression11 (44%)15 (60%)Level of supportLow or none5 (20%)5 (20%)Medium13 (52%)12 (48%)High7 (28%)8 (32%)Ethnicity*Asian/PI1 (4%)9 (36%)African American6 (24%)7 (28%)Latino/Hispanic8 (32%)2 (8%)Non-Hispanic White10 (40%)7 (28%)Improved*16 (64%)8 (32%)Age (mean, SD)38 (14.64)42 (16.32)*Treatment groups differ, p < .05 ________________ Hypothesis Testing What are the independent (including control) variables for the multivariate analysis, and their level of measurement? Treatment Group Categorical Depression history of client Categorical Age Continuous Parents history of depression Categorical Asian or Pacific Islander Dummy variable, categorical African American Dummy variable, categorical Latino Dummy variable, categorical High social support Dummy variable, categorical Medium social support Dummy variable, categorical What is the dependent variable and level of measurement? Improvemed (vs. not) in mood and symptoms -- Categorical State the null and alternate hypotheses for the overall model fit. Null: Improvement in mood and symptoms is not related to the independent variables as a group. Or, the model does not predict the likelihood of improvement in mood and symptoms. Alternative: The overall model is predictive of the likelihood of improvement. State the null and alternative hypothesis for the main independent variable. Null: Type of treatment will not differ in the extent of improvement of mood and symptoms. Alternate: Group therapy will result a greater likelihood of improvement in mood and symptoms than individual psychotherapy, controlling for age, ethnicity, social support, and client and parental depression history. Why is logistic regression the correct statistical procedure? Dependent variable is categorical and there is more than one independent variable. Review SPSS output. (SPSS commands: Analyze; Regression; Binary logistic; Options: Hosmer Lemeshow Goodness-of-Fit) Hosmer and Lemeshow TestStepChi-squaredfSig.19.3248.316 Variables in the EquationBS.E.WalddfSig.Exp(B)Step 1treatment_group2.032.8635.5451.0197.629dep_hx.475.731.4231.5151.609Age.001.024.0021.9691.001Parent-.801.7421.1651.280.449Asian_PI1.4671.1121.7411.1874.338Afr_amer.040.959.0021.9671.041Latino.779.969.6471.4212.180High_support-2.3961.1094.6631.031.091med_support-2.6041.0975.6351.018.074Constant.6221.647.1431.7061.863 Summarize results and decision to reject or not reject null. The Hosmer Lemeshow test showed that the model predicts the observed improvements in mood and symptoms, X2 (8) = 9.324, p = .316. We cannot reject the null that says the model predicts the observed dependent variable values. Group therapy results in a higher likelihood than individual psychotherapy of improvements in mood and symptoms, Exp(B) = 7.629, p = .019. Group therapy participants were over seven times as likely as those receiving individual psychotherapy to show improvements in mood and symptoms, controlling for age, ethnicity, social support, and client and parental depression history. High levels of support (ExpB = .091, p = .031) and medium levels of support (ExpB = .074, p = .018) show lower likelihood of improvements than low levels of support, controlling for all other variables. Neither variable had as much impact in the likelihood of improvement as group psychotherapy. None of the other independent variables had significant effects on the likelihood of improvement. Discussion and Limitations Include: Address original research questions or hypothesis Meaning and implications (including relevance to Transcultural perspective) Limitations of study Recommendations for further research Our hypothesis was supported that group therapy had a greater likelihood of improving mood and symptoms, compared to participants in individual psychotherapy. Six months later group psychotherapy showed far greater reductions in depression symptoms. The use of time-limited group therapy seems to be a reasonable alternative to individual psychotherapy. It remains to be seen, however, whether the gains made in group psychotherapy hold after six months compared to a longer trial of individual psychotherapy. Interestingly, higher and medium levels of social support, compared to low level, had a lower likelihood of improvements. One might expect higher levels of social support to be associated with more improvement. It is unclear why these results were obtained, and further research is required to understand the role of social support in the improvement of symptoms of depression. One possible explanation is the small sample of low level cases resulting in an incomplete test of the role of social support. While ethnicity did not play a role in the outcome, the sample size did not allow for adequate comparison among the ethnicity groups. Further research replicating the experimental design with a larger sample would be required. A limitation of this study is the possibility of treatment diffusion that cannot be controlled. For example, it is possible that those in individual psychotherapy received some benefit of group interaction outside of the treatment environment, such as obtaining informal advice or support from friends or family. If this occurred, the impact of group therapy could have been understated. Treatment diffusion could also have worked in favor of group therapy, if group members received individualized advice, information, or supportive interventions outside of the group environment. If this occurred, the importance of the effects of individual treatment would have been understated. Collecting data measuring the extent to which treatment diffusion occurred would help explain this. These issues point to the need in clinical treatment experiments to specify the type of therapy being used and provide training to clinicians to ensure that the treatments being compared are implemented with fidelity.      PAGE \* MERGEFORMAT 5 IJXYZ[nyz  V B C b = o q g o ʿ}h hp56hphVh{h3h5Eh7Xh"dhZ hPMh:hj&hh%hEhGch5CJ \aJ hw5CJ \aJ h:h:CJ aJ h:h:5CJ \aJ h:5CJ \aJ h:h:02<IJ[z{  $IfgdI0$a$gdEgdE   & B w x uv%&QR\]?@qr  !9YZ\^aceĿȭhq-hEhM=h&Irh&Irh&Ir6 h&Ir6h]hc5 hM_5 h]5h]hj h{CJ hK+CJh{3h CJh{3h[CJh[h{3h[5CJhEhI0:  & B Z w }wwww$Ifkd$$Ifl440 #x  t0644 laf4yt{3w x  {rr $Ifgdl$If}kd$$Ifl0 #x  t0644 layt{3 {{$If}kd$$Ifl0 #x  t0644 layt{3 {{{$If}kd$$Ifl0 #x  t0644 layt{3Gu{{{$If}kd)$$Ifl0 #x  t0644 layt{3uv{rr $Ifgd4c$If}kd$$Ifl0 #x  t0644 layt{3%{rr $Ifgd4c$If}kd9$$Ifl0 #x  t0644 layt{3%&3{rrr $Ifgdw$If}kd$$Ifl0 #x  t0644 layt{3x{rrr $Ifgdw$If}kdI$$Ifl0 #x  t0644 layt{3!.?{rrr $Ifgdw$If}kd$$Ifl0 #x  t0644 layt{3?@!:ymmmm $$Ifa$gdk$If}kdY$$Ifl0 #x  t0644 layt{3 :;Xaj $$Ifa$gdk$Ifgkd$$IflF2 d 2 @  t6    44 laytkeg #%56=>EG/CMlovBKWm¾ hF5hA>h;x5hA>hC3/5h)Rhh="hq_hEX hq_5 h{5 hM=hj hM=6hY#hFhlaOhvhM=h?hU}h&IrhE=jk $$Ifa$gdk$Ifgkd[$$IflF2 d 2 @ t6    44 laytk $$Ifa$gdk$Ifgkd$$IflF2 d 2 @ t6    44 laytk $$Ifa$gdk $$Ifa$gdkgkd$$IflF2 d 2 @ t6    44 laytk $$Ifa$gdk $$Ifa$gdkgkdu$$IflF2 d 2 @ t6    44 laytk $$Ifa$gdk $$Ifa$gdkgkd$$IflF2 d 2 @ t6    44 laytk    $$Ifa$gdk$Ifgkd1$$IflF2 d 2 @ t6    44 laytk  $ $$Ifa$gdk $$Ifa$gdkgkd$$IflF2 d 2 @ t6    44 laytk$%6>F $$Ifa$gdk $$Ifa$gdkgkd$$IflF2 d 2 @ t6    44 laytkFGW_f $$Ifa$gdk $$Ifa$gdkgkdK $$IflF2 d 2 @ t6    44 laytkfgz $$Ifa$gdk $$Ifa$gdkgkd $$IflF2 d 2 @ t6    44 laytk $$Ifa$gdk $IfgdlaOgkd $$IflF2 d 2 @ t6    44 laytk $$Ifa$gdk $IfgdlaOgkde $$IflF2 d 2 @ t6    44 laytkN} & FgdFgdq_gkd $$IflF2 d 2 @  t6    44 laytk B|}stmnHIJRTm $d@$Ifa$ & Fgd7pgd{3gd{3gdA}gdA}gdKYgdq_gdq_ & FgdFm{|}>PrstmnFGHQSTl1羺穡癔Ύ}ok}ooh7hKYCJOJQJ^JaJ hKY5CJOJQJ\^JaJ hKYaJ hcB15hcB1hRhh{35 h5 hODhKY hOD5hFhA>hF5hKhK5hKYhA>hlu5hA>h;x5 hK5h{3hA}hA>hKY5hA>h9E5&mns~xxx $d@$Ifa$ d@$Ifrkd/ $$If 44`^  0`24 af4p 3* $d@$Ifa$ d@$Ifkd $$If 44\w ` (0`24 af4p((kda $$If 4\w ` (0`24 af4p( $d@$Ifa$|vv$Ifrkd$$If 44""  0"24 af4p $d@$Ifa$gdKY    "$)/015:Ff$IfFf $d@$Ifa$ d@$IfFf $d@$Ifa$12STqrs#(*+SX 23eghv{5 x | ѾѹѹѴ|upkf hC5 hM 5 ht5 h(?5>* h(?56 h(?5 hV?56 hV?5 hOD5 hf<5 h'56 hz=5 h 5 h'5 h{35 hmt5 hEOj5 h#5H* h#56 h#5h{3hKY hKYaJhKYCJOJQJ^JaJhM CJOJQJ^JaJ(:?DFKQRSZ`ekmrwxyFf'Ff}# d@$If$IfFf $d@$Ifa$ Ff0Ffo, $d@$Ifa$ d@$If$If)06<>CHIJSX^cejpqrtgdKYdgdKYFf^>Ff9 d@$If$IfFfa5 $d@$Ifa$t :!O!t!s#p%U&@*B*D*E*gd{3gd &$d%d&d'dNOPQgd2 & Fgd1;gd1;&$d%d&d'dNOPQgdHvgd{3| t!!!!n"r#t#########G$Y$t$$o%p%U&V&k&q&t&''''''|(()R)e)?*@*A*B*C*E*F*H*ȾȶױעעᢘғҎ܆~z~zhG.jhG.Uh hC5 h S5 h5 h5 h(&5 hZ5 h25 hFt/5 h'5hTQhK+5hTQhTQ56 hTQ5 hY5 h=5n5 hK+5 hC5 h5 h 5h1;h{3 h(?h' hM 50E*G*H*J*K*M*N*j*k*l*m*n*gd $a$gd{3 H*I*K*L*N*O*f*g*h*i*k*l*m*n*h hC5hC6Rh*mHnHuh{3jh{3UhG.jhG.U 21h:pd/ =!"#$% $$If!vh#vx #v:V l44 t065x 5f4yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3$$If!vh#vx #v:V l t065x 5yt{3x$$If!vh#v #v2 #v@ :V l t65 52 5@ / ;/ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytk\$$If!vh#v #v2 #v@ :V l t65 52 5@ ytkj$$If!vh#v #v2 #v@ :V l t65 52 5@ / ytk$$If !vh#v`:V  44  0`,5^/ 244 af4p i$$If !vh#v#v#v#v:V  44 (0`,5555/ / /  / / 244 af4p(f$$If !vh#v#v#v#v:V  4 (0`,5555/ / /  / / 244 af4p($$If !vh#v":V  44  0",5"/ 244 af4p $$If !vh#vN#v#v#v#v:V  44 P0",5M5555/ / /  / / /  / / 244 af4pPkd$$If 44ִN 3H]r"M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / /  / / /  / / 244 af4pPkd$$If 44ִN 3H]r"`M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkd$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkd$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkd!$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkd&$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkd}*$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkd.$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkdo3$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  44 P0"+,5M5555/ / / / / / / 244 af4pPkd7$$If 44ִN 3H]r" M P0"    24 af4pP$$If !vh#vN#v#v#v#v:V  4 P0"+,5M5555/ / / /  / / /  / 244 af4pPkda<$$If 4ִN 3H]r" M P0"    24 af4pP666666666vvvvvvvvv666666>6666666666666666666666666666666666666666666666666hH6666666666666666666666666666666666666666666666666666666666666666662 0@P`p2( 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p8XV~8XV~OJQJ_HmH nH sH tH J`J RuNormal$1$CJOJQJ_HmH sH tH R@R * Heading 1 F<@&KH PJ\^JaJ N@N * Heading 2 & F<@&\]^JaJ`@` * Heading 3($ & Fddh1$@&[$\$ \^JaJL@L * Heading 4 & F<@& \^JaJR@R * Heading 5 & F<@&PJ\]^JaJDA`D Default Paragraph FontRi@R 0 Table Normal4 l4a (k ( 0No List TT RuHeading 1 CharCJKH OJPJQJ\^JaJ JJ Heading 2 CharCJOJQJ\]aJLL *Heading 3 CharCJOJQJ\^JaJH!H Heading 4 CharCJOJQJ\aJ424 Z,Style1 FOC T0APAa:VjZ dCJOJQJ^JaJ5NQN Heading 5 CharCJOJPJQJ\]aJbobb :Default 7$8$H$-B*CJOJQJ^J_HaJmH phsH tH tst  Table Grid7:V0 d44 {30Header  H$BB {30 Header CharCJOJQJ^JaJ4 @4 {30Footer  H$BB {30 Footer CharCJOJQJ^JaJPK!pO[Content_Types].xmlj0Eжr(΢]yl#!MB;.n̨̽\A1&ҫ QWKvUbOX#&1`RT9<l#$>r `С-;c=1g'}ʅ$I1Ê9cY<;*v7'aE\h>=,*8;*4?±ԉoAߤ>82*<")QHxK |]Zz)ӁMSm@\&>!7;ɱʋ3װ1OC5VD Xa?p S4[NS28;Y[꫙,T1|n;+/ʕj\\,E:! t4.T̡ e1 }; [z^pl@ok0e g@GGHPXNT,مde|*YdT\Y䀰+(T7$ow2缂#G֛ʥ?q NK-/M,WgxFV/FQⷶO&ecx\QLW@H!+{[|{!KAi `cm2iU|Y+ ި [[vxrNE3pmR =Y04,!&0+WC܃@oOS2'Sٮ05$ɤ]pm3Ft GɄ-!y"ӉV . `עv,O.%вKasSƭvMz`3{9+e@eՔLy7W_XtlPK! ѐ'theme/theme/_rels/themeManager.xml.relsM 0wooӺ&݈Э5 6?$Q ,.aic21h:qm@RN;d`o7gK(M&$R(.1r'JЊT8V"AȻHu}|$b{P8g/]QAsم(#L[PK-!pO[Content_Types].xmlPK-!֧6 -_rels/.relsPK-!kytheme/theme/themeManager.xmlPK-!!Z!theme/theme/theme1.xmlPK-! ѐ'( theme/theme/_rels/themeManager.xml.relsPK]# n"  ***- em1| H*n*%4:?A w u%?:j $Ffm:tE*n* !"#$&'()*+,-./012356789;<=>@ $&-!8@0( ltnc ltnc  B S  ?26%x~v~&2BL178@TZ_g~ y(#(%)Z^A"B"B"D"E"E"G"H"J"K"M"N"i"l"o"%0 " # p | UZ~    B"D"E"G"H"J"K"M"N"l"o":::::::::::::: ]ZH& 8-EaSGٔqV{PDWR"OrrRj^ }QxF>ye~) ^`OJQJo( 8^8`OJQJo( ^`OJQJo(o  p^ `OJQJo(  @ ^ `OJQJo( x^x`OJQJo( H^H`OJQJo(o ^`OJQJo( ^`OJQJo(`^`o(.^`o(.@p^@`p56CJOJQJo(. h ^ `o() 8 ^ `o(()^`o(()p^p`o(()@^@`o(()x^`o(()^`.^`.pL^p`L.@ ^@ `.^`.L^`L.^`.^`.PL^P`L.^`.^`.pL^p`L.@ ^@ `.^`.L^`L.^`.^`.PL^P`L.^`.^`.pL^p`L.@ ^@ `.^`.L^`L.^`.^`.PL^P`L.^`.^`.pL^p`L.@ ^@ `.^`.L^`L.^`.^`.PL^P`L.^`.^`.^`.p^p`)@ ^@ `()^`()^`()^`()^`()h^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hH^`.^`.pL^p`L.@ ^@ `.^`.L^`L.^`.^`.PL^P`L. ^`OJQJo(^`OJQJ^Jo(o p^p`OJQJo( @ ^@ `OJQJo(^`OJQJ^Jo(o ^`OJQJo( ^`OJQJo(^`OJQJ^Jo(o P^P`OJQJo(Sy-E-Eg>Gg>Gg>GF>y-Eg>G-E-EZH& ZH& ZH& V{PWWOrr}Qx (        4EJ        b        -                          g                 dt.P^N\hxb *C Z rM=W+wu| C*)R-j ="K"$s$x$9%(&C)3+K+Z,q-G.C3/Ft/I0cB1{3q4b6d8A99:1;z=jE=A>V?? @5EE IuJ0~JKPM+NlaOZ9PTQC6RTJWEXY]S^c^M_q_4c3 d"dGchEOj=5n7p r&IrmrzrtluHvsvA}<>Fk)$T@l'1j\6:rj&ERwD(?mt"{{rAaO#[vU}tE SYq9ET]7:jo>V-'2<ZaqY#FO[zd9WE $M 3^{[nw%{07XcE[Ru4<&>>A T[;xpAF &ODC QKY\7~9B"D"@@zn"P@UnknownGTimes New Roman5Symbol3 Arial7Calibri? Courier New;WingdingsACambria Math"1hugug զ&>&>!x201"1"3QHP $P:2!xx ,ecohen Chico Cat8          Oh+'0  8 D P \hpx'ecohen Normal.dotm Chico Cat2Microsoft Macintosh Word@F#@j@^K@^K& ՜.+,0 hp|  ' >1"  Title  !"#$%&'()*+,-./0123456789:;<=>?@ABDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcefghijklmnopqrstuvwxyz{|}~Root Entry FbsMData C@1TabledAWordDocument4SummaryInformation(DocumentSummaryInformation8CompObj` F Microsoft Word 97-2004 DocumentNB6WWord.Document.8