../.V.t. .5... .. V. . ._ 4%.... an... Vmaimxmuwma . .V . .. V. . $3.3: V .VV.. ..‘~.\1..V.V..\fil . V . V.. .umguufivigvflf! ..V V . .1... .TrIV .w... .5... l 8...- .V... a... 3%. . . .. 1:9. wuflnwéf. m . mvakm fifimmhwwwwm. ., V . 2%,.ficw. A. .1 . .29 Egg... V. L L 3.. . H... V V .. V 5%.. .1...qu.... fit... .. _ _ .V in”. kmmw... n 3.3....wfaV3. V, 5... fifiw . Z V 2.. . P. Vida. ,1 ..V 1.4 . .. 1.. 8%. .9. . r . “V... V . . .. V .. ha. I. 1\ . 1%“. V Ma. r {lutnx g ml; )7 “Ham V. V \ . Ar .3 Q . « 20.1 III!!! gs»... .etetV .mm.3e:(e....rh:r. .. \YK..\~IL a. IQICIVJI. 3% i 10.1.3. fflaaxwi.sau.._s.l . . .V . V... V, in: V. a A“... : Nani x..- . it; \ 1.5%)...» 4. . 5.3!... . . L... 5.313. 3 w. 1 7y I 119V; .. 1... V . L. 31.. 3| . .. it: it; _ V. 1...... . . a... mum???.mfi.§fi.;fi.3 ... . . 5 ._...,. .n en fivfifijfiShirt:r...z...rcux...§.e¥..f, . tunmofxflhflufi fins... .V Vfimfigwhnhn?§fi§mmmfir V... L...m%h$ .3? . H . .2 . V ...znan§..w.wfi .-....V.h..r...-»¢ii.i .Nflflf .1....E3fi.. H... . a. 1.3. V. , .r V V a 1 . . . . .. . .. J A. . ( ( . ‘Axtfli'l! “V 3.... . V .89». 5.. ... .3 9V .. Sfitfihqki. VK. Whaiutbflofll. filliitur. z. z a. as .35..an ._ . In E. .1? ha .. b... 5.? .i...11 -V V V. V. _ .Vwfiwa wfifiwwufsfififiw a... (gilt-5|, 3 h Q ‘s: l-‘rf‘g s‘ug’:- . 1. 5s 1. LZII! EiI zOhSUhn‘hntlfait-Nn Ix. . . . \ V 1 , EV F. {tinisiza‘ 35 .5 l . . a .V .. a. . \ . a :13‘. a}? h} Q. 7 km); V. «L. V.. fiu. V . . V . i 63.....532 1995.. Cavity 5...}- 3.0).! I \ 1 .f . x. V 903...}. .\ . . 1...... i. . x . V 1 i... «1.1 I. - 2 [‘SIQ .. 0%...1ll.§ V 5......1. . r . u“. :2: 61...... . v\.U>..whme. V. L .0. .999 .V. r n... . . . ,. . . til. 2.2.... .39.“ EOE-Hat“ WIQH~XIBQI~WH~§ .(AVEI‘ ‘ V V . , .. a V . V V . r . V V y . . DINA” .9311... x. S M. . VV . h! C 103. . . LNLN . .VI4991. than-i. . . V ~ as? L.fiHQA._Wau )qh.) h4|aleOOLnUAaLLNQI 3.5.5 \ 4A,... V _ . . v e V a. 03 a. V9 3.... , i .5...) ahut . I... . . ..4\EI9.IAI...:§L§ .2. hrs... M087. £nui§hl£é~hfll3rf .u. t r: . u a: .. x V doWfid...‘ a... . Rm. . go.) 38.“. MN. . . .V. 55. Levis.vu.rflnltntt¢r¢V he.“ ‘ 4‘..." $5....uifiut .Lst..btl.\..tl..V.‘l..l..§\||b|.fl . . 7...... s z 2.. V. . ,. V .3. .q s, .z . E... .3..- . £15. 212...... p #1. 3.." I. u‘ 3 21? . . . . V. T . .IALIIII...1§AV.}V 3.117) I A.%lfiih\hflh.\ qulhnh...\\|bnl: llldlll. . .. . it”)... 1.. d5»... V .dsvg up... . . V v. . r . c: Igi \ adufiwv .V. 25;.“ ‘5”..th ’nflfitteivlflndiinhflkf|lir . . V V . . . V .V u .. ..\. I £4. E . . «.75. are uni 111).... 709317.39. .fl . . 5:69. RH"... :\.h\£.v‘. “Mk.“fllaaflflg‘l) in}!!! {I‘d-“Luau.“ . “Ill-6 ‘1.“ . V Vlkk. . Li - L. x . .VI \..~(f.(l§..t$!£u~l. V .Kkflmz. in“ I. .35.. 1 HR... . V . V zttt1\.\\ .Uwrdbuthn . 5s t\~IIl.|7 EL! .. . . 36‘ . . I) Vt... 3x 7 V ixliotnx.mr . x A. Y E .‘2‘59 3“ V 3 193%? V Elli/II ‘ .7»...— Aflflwutlil stat}? 3!?!- 5... ‘? tlx..b.u‘\«l. lilet-(onVszftr‘ixl’i It}. i A“)?! I.“ 5.. . . L :1.» :1. .5. V . .. . \.~i~ t it. Q‘ssz‘iti basil}; \|€\¥ .. :3. u. l2 :5. ~ \« ........... 44;.-11! 51.5 n . , g f: .3 Mfl. J: .. gag“... 992K. it..t£l.ati.lhnw!y>hz Viv t 1. 5:1 ~3.IKP“L\% tQ i.\.....!s.\ 5m .iailx u~§339l ‘35.]!!! A... \ I£§.11, . O... x Wynn-5. \s 3:3th MN“ {.1 1>\V Hutsfispfitfil. It ’Olohfi. . ‘al I . All) “.1: {I a .U 1 Juli. .Iuss gunwoiwumw‘tfiu . \ 11.“... 00133.11. 95):; }sn_n-.zv.h8¢r(h$dr\f:fl}.h . V . 1V . . V.. \ V... I c. I; $\\‘1€ .ia S4. \ m $3.3. . a... unidwfifluugtb .éngiu. agicttsfinfixcdbtc sin" VV . l V $8.1?! is) i :12! ..n1|I.-..lv1...¢“~lil( 1.. . a; Q .mmmz. . .. 5.2.3.... 5.2.2. :1 . . ~ .~ '3. z X? . Ill|§in..\sv).\k\l.v}1.I-. . x213. ‘fiflfi. 9 a 5:: s. It 3%.“... . V. . .2 x. . h. .\ a» .V . has... #1:..mmuulxflqzztz. xix-is.ti...?uiu.10hu.h.uv.1x1 \ (a 5.3:: V )J at. Iiihckw‘iuflflh u .0” a..‘!4¢~l\’.-¢.1\19|\¢(tl‘€‘si.iidhbi \SIIt)ll.-.Iizll H... V . . VV 1 in... 5...... e6... :5. .7..- .. vir' ‘h‘ . \‘15‘...IAI"|$.| \ 121-..... fix”. . . 3. its-5.1:!!! 4L .1: .5... .aglssrrl}. 3).?597? :9... .152 §\.r .3. a V . x 3: . 1.2;}: 51.)). 1...)...runfifl. 53333325. huh: .. Z (9.21:)... I V..¢y.(rl.\\..t\3. 1.2 x-.. s hum. (. .2; #33135, it . . r71. Ilf To. a. VV Pct—u.- j} {P >1»! .5 i; , L2. . .1: .. XI . ‘5. .. . .. TT...‘ «Ration?! .9! V. V .fln . bumnzuqfl . .. .. . x.» lung. . . V 93.1.. .V “5.9.2: Rag . 1 c. V A. Aist. . V %I§%. 51.....‘0dfia L :-21:51.45:imufi...’90£htha~~t«»k-.fo?l\\§faA.‘\Al..‘\.((urL 7.. z 3‘ . . . 60‘s. .plu- 5‘. Exit} . . . ...r. .3125: 93 5%.. ..V ..S..A$uun. 5539).“ . m . x: 1.311;??? “33‘ .2 \\4ls.u¢€\(l 1.7.5.59?! t:...~.i«l«9\\\..¢\...tun\t5 ding??? y 53.3.: ‘53.! 3 III: 7.1 ~..I..r1\1.«\\~ . . . y .. $3.25.. Lu. 27.5.5...) :51 79.23%... 23?. . V V... in”)? 1(..\‘C¢.§7u4.ul}.ftiitl filizmfité .vxn . . . . lot\\\)~... . . .i h.s .& .;§!§9§ 3.: n"....§4\t\§.§\tba \t;\§l€l. \ .c 555:... #12 5. .3351 1£|§§t$422tfg V s}. ztttt’fr 4. r. .:.\l.!tl:f.§V .a!|31hs\sk€tt 2‘33) V . 5: ‘(t t2. 55.501...) . V fihhtg:$§x .r... “liar”. t..§~.£..1|(l§. 1:5..‘1111VIA yr. .(. \\I0\0¢1‘\¢anunti.l\. l S i t I . 9.. V. J . . 1.3! ‘hkAalllA‘sxk {1):}...(‘ILSO‘DM ft .JK.‘ .II'.‘I\34VII.‘. Wu“ drank, 1.3 X .1. I .3730}... .IL.\\ 5.71%: 1Vi\\§..2‘¢11 «trait. {ILA (\l\1\ Ix. . .. .. .mfinufrzs .. 5.51.31. 52.2.}: . h! ..ytl.,i2.2...§..$fs9 33...»... $341 131. . Z5 5.5 L. Lt .. 5.4? anV ~; #4 2... L5... .. . 377.11 153:?! £95... . 24 fizznflvnul- 1 3. It. ....-§z-§a2 .. vfiun . .3331... s V V Karisi... It. .2. 3‘34““. .393...“- ....r, m... .i V V .< v‘ .\rl. x g.Vcs’w-¢.ua‘h ....V..¥Vm:....éw.., is. J ’~I¢o’.\l\h: i; g 3.35;“ u... . if?! 31:115.“. 1.573111%... :...7:~Z(3 5"AK .33] .51.,12‘.‘ linki‘ii all“? . . .55.!I7I4‘n Ié‘ Zir’fi ‘ 1.4!. Eifiuigiz; 11:3. .5. i! N \lV 1.3.1. V. .. .Km. :2 1.2. I s. '57-». v 5.1.3: 13:13.1... “.7. yin-5%.. 1?: 3.3%.)...» . 3 z :s u V .6. 1.1.3; 2.8: £3314. )‘4’. '9 :\~\ .\ I. ‘95:... his... ll‘liizt "1|”- ‘.» :1 _ LIBRARY Michigan Stat: University This is to certify that the thesis entitled AN ANALYSIS OF RESTRICTED WORD ASSOCIATIONS OF ELDERLY PERSONS presented by Carolyn Marie Hagey has been accepted towards fulfillment of the requirements for Ph. D. degeehlAudiology & Speech Sciences gig/flew 'or professor DateM 0-7639 © Copyright by CAROLYN MARIE HAGEY 1978 AN ANALYSIS OF RESTRICTED WORD ASSOCIATIONS OF ELDERLY PERSONS BY Carolyn M. Hagey A DISSERTATION Submitted to Michigan State University in palfiLial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Audiology and Speech Sciences 1978 We ABSTRACT AN ANALYSIS OF RESTRICTED WORD ASSOCIATIONS OF ELDERLY PERSONS BY Carolyn Marie Hagey 'The investigation and rehabilitation of semantic confusions found among elderly aphasic patients has been limitxed. in scope due to the comparatively meager knowledge of restricted word association responses elicited by nonpatho- logicxil ("normal") elderly persons. In order to distinguish between linguistic changes resulting from neuropathological condijzicnis and those changes (if any) caused by the aging procefigs, it is first necessary to have a thorough understanding of "ncnflnalf' language usage among the elderly population. Thus, tine purpose of this investigation was: 1, to provide data on four restricted word association tasks presented to a group of nonpathological elderly subjects, 2. to investigate the effects of age, sex and years of formal scflnooling on the restricted word association responses eliéited by tflue elderly subjects. One 1Tundred thirty-two nonpathological elderly persons, who passed a visual, hearing and mental ability screening procedure, Served as subjects. All subjects were white, Carolyn Marie Hagey native loorn American English speakers. Subjects were dividend into two age groups (65 to 74 years, 75 years and above) ; two sex groups ( males, females); and two educational groups (less than high school education, equal to or greater than high school education). A synchronized audio-Visual presentation of 25 nouns was used to elicit responses for four restricted word assoc- iation tasks. The four task conditions consisted of two logical tasks (superordination, similarity) and two infra— logical tasks (location, part). The same 25 nouns were presented on each task. However, the ordinal presentation of the noun stimuli varied from task to task based on the "easiest" to the "least easiest" stimulus item on each task. Subjects were seen individually for approximately 40 ndJuutes. Each of the four tasks was presented using the following format: 1. subject read task instructions 2. three practice items were presented 3. questions of clarification were answered 4. noun stimuli (n = 25) were presented . Respcnlses were graphically and electronically recorded. Data were analyzed employing descriptive and infereurtial statistics. Descriptive statistical analyses were used t0: 1. determine the common responses and corresponding frequencies (3f occurrence elicited by 85 percent of the elderly pOEn11ati0n sampled on the four tasks. Carolyn Marie Hagey 2. determine the primary responses and corresponding frequencies of occurrence elicited by the elderly pOpulation sampled on the four tasks. 3. determine the diversity scores for each lexical stimulus item on the four tasks. Results indicate that the infralogical location task had the highest amount of response agreement, whereas the logical similarity task had the highest amount of response vari- ability. This pattern was consistent irrespective of the age, sex or educational level of the elderly subjects. Interpretation of three two-way, fixed effects analyses of variance yielded the following results: 1. Overall primary response agreement was unaffected by the age of the elderly subject. Results indicate that restricted word associations remain relatively stable for elderly persons aged 65 years and older. 2. Female elderly subjects had significantly higher response agreement than their male counterparts. 3. The number of formal years of schooling influenced the response agreement obtained by elderly subjects. Persons with more years of formal schooling exhibited higher response variability than persons with less years of formal schooling. To my Grandparents, Dad and Mom: the stems from which I blossom! To Dr. John W. Black and Dean Frank Hale: for providing the water! ii ACKNOWLEDGEMENTS Research studies are never solely due to the endeavors of the primary investigator. As with many studies that have preceded this one, I find myself indebted to many people. A sincere thanks to Mark Greenwald for his assistance and moral support throughout the research study. Recognition is also given to Dr. June Sears and Dr. Jane Oyer for their assistance in locating the elderly population investigated, to Mr. Howard Doss, for extending and re—extending the loan period on the special equipment needed to conduct the research project, and to Suwatana, for her assistance in computer programming and interpretation of the output data. I am also greatful to my typist, Sharon Nowland, who spent personal time helping me meet my deadlines. Finally, to my committee members — Dr. Leo Deal, Dn Herbert Oyer, Dr. Linda Gillum and, especially, Dr. Abrahm Barch _ who helped my research idea become a reality! TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . Vii LIST OF FIGURES . . . . . . . . . . . . . . . . . . ix Chapter I. STATEMENT OF THE PROBLEM . . . . . . . . . . 1 Introduction . . . . . . . . . . . . . 1 Purpose of the Study . . . . . . . . . . . ll Terminology . . . . . . . . . . . . . . . 11 Research Questions . . . . . . . 12 Organization of the Dissertation . . . . . 14 II. REVIEW OF RELATED LITERATURE . . . . . . . . 16 Introduction . . . . . . . 16 Historical and Current Conceptualizations of Adult Cognition . . . . . . . . . . 18 Summary (Part 1) . . . . . . . . . . 27 The Relationship of Semantics to Word Association . . . . . . 28 Effects of Age on Responses Obtained on Word Association Tasks . . . . . . 30 Effects of Sex on Responses Obtained on Word Association Tasks . . . . . 33 Effects of Educational Level on Responses Obtained on Word Association Tasks. . . 34 Effects of Mode of Stimulus Presentation on Word Association Responses . . . . . 35 Procedures Employed to Analyze Data From Word Association Tests . . . . . . . . 37 Summary (Part 2) . . . . . . . . . . . . . 39 iv Chapter III. RESEARCH METHODOLOGY . . . . . . . . . . Subjects . . . . . . . . . . . . Screening Procedures Used For Subject Selection . . . . . . . . . Stimuli . . . . . . . . . Generation of Task Stimuli . . . . . . Test Environments . . . . . . . . . . Equipment . . . . . . . . . . . . . . Test Procedure . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . IV. RESULTS AND DISCUSSION . . . . . . . . . Results . . . . . . . . . . . . . . Discussion . . . . . . . . . . V. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Summary . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . Recommendations . . . . . . . . . . BIBLIOGRAPHY . . . .. . . .. . . .. . . .. APPENDICES Appendix A. Questions Used on Mental Status Questionnaire . . . . . . . . . . . . B. General and Individual Task Instructions . . . . . . . . . . . C. Sample of Lettering Used to Create Stimuli Slides . . . . . . . . . . D. .Answer Form Showing Order of Stimuli Presentation on the Four Tasks . . . E. ‘Alphabetized Primary and Common Responses and Associated Frequency Distribution for Superordination Task . . . . . . o Page 40 40 41 45 49 52 54 58 60 60 96 110 110 112 114 116 124 125 130 131 135 Appendix . Page F. Alphabetized Primary and Common Responses and Associated Frequency Distribution for Similarity Task . . . . . . . . . . 147 G. Alphabetized Primary and Common Responses and Associated Frequency Distribution for Location Task . . . . . . . . . . . 162 It: Alphabetized Primary and Common Responses and Associated Frequency Distribution for Part Task . . . . . . . . . . . . . 172 vi Table l. 2. 3. Noise Levels for lOOOHz, 4. Dictionaries 5. Norms . . . . 6. Subjects' 7. Primary Responses of Occurrence (f) 8. Primary Responses of Occurrence (f) 9. Primary Responses of Occurrence (f) 10. Primary Responses of Occurrence (f) 11. LIST OF TABLES Estimated Incidence and Expected New Cases of Stroke Per 1,000 Persons Number of Errors Associated With the Degree of Organic Brain Dysfunctions as Determined by the Mental Status Questionnaire Comparison Between the Allowable Ambient 2000Hz and 4000Hz and the Actual Ambient Noise Levels Observed in the Three Test Environments . . . . Difference in the Number of Words Classified Under Three Form Classes Based on the First Part— —Of— —Speech Entry of Two American English List of the Twenty—five Test Stimulus Items Selected From the Original Two Hundred Test Stimuli of the Michigan Restricted Association List of the Eight Groups Included in the CrosstabulatiOn Analysis as a Function of the Age, Sex and Education . Mean Frequencies of Occurrence of the Primary Responses Under Four Task Conditions and for and for and for and for Corresponding Frequencies Superordination Task Corresponding Frequencies Similarity Task Corresponding Frequencies Location Task Corresponding Frequencies Part Task Page 42 44 47 48 62 63 64 65 66 68 Table 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Diversity Scores for Each Lexical Stimulus Item Under the Four Task Conditions . . . . . Summary of Mean Diversity Scores as a Function of Age, Education and Sex . . . . . Summary of Correlational Values of the Diversity Scores of 25 Lexical Stimuli as a Function of Sge, Sex and Education . . . . . Cumulative Frequency: Superordination . . . Cumulative Frequency: Similarity . . . . . Cumulative Frequency: Location . . . . . . Cumulative Frequency: Part . . . . . . . . Summary of Mean Values for Primary Responses Given as a Function of Age, Sex and Education . . . . . . . . . . . . . . . . . . Two Way Fixed Effects (Age X Task) Analysis of Variance with Repeated Measuses and Equal Group Size for the Number of Primary Responses Elicited by Subjects . . . . . . Planned Comparison Post Hoc Analysis of the Task Main Effect For All Subjects . . . . . Two Way Fixed Effects (Sex x Task) Analysis of Variance with Repeated Measures and Unequal Group Size for the Number of Primary Responses Elicited by Subjects . . . . . . . Planned Comparison Post Hoc Analysis of the Task Main Effect For All Subjects . . . . . Two Way Fixed Effects (Education x Task) Analysis of Variance with Repeated Measures and Unequal Group Size for the Number of Primary Responses Elicited By Subjects . . Primary Responses Identical to Those Appearing in the Michigan Restricted Association Norms (Riegel, 1965) . . . . .. viii Page 70 72 78 79 80 80 81 82 84 87 89 91 93 100 LIST OF FIGURES Figure Page 1. Demonstration of the Insertion of the Projector Command and the Tape Stop Command Signals for Obtaining a Synchronized Audio— Visual Presentation of the Test Stimulus Item 53 2. Physical Arrangement of the Equipment and Subject in the Test Environment . . . . . . . 56 l 3. Mean Diversity Scores for Superordination ‘ Task as a Function of Age, Sex and ‘ Education . . . . . . . . . . . . . . . . . 73 4. Mean Diversity Scores for Similarity Task 'as a Function of Age, Sex and Education . . . 74 5. Mean Diversity Scores for Location Task as a Function of Age, Sex and Education . . . 75 6. Mean Diversity Scores for Part Task as a Function of Age, Sex and Education . . . . . 76 7. Mean Primary Response Scores for Male and Female Subjects Under the Four Task Conditions 90 8. Interaction Between Education and Task Condition . . . . . . . . . . . . . . . . . . 95 ix CHAPTER I STATEMENT OF THE PROBLEM Introduction The 1971 White House Conference on Aging made numer- ous disciplines of social science cognizant of the paucity of research which examined the biological, social and cogni- tive behaviors characteristic of the elderly population. For purposes of the present investigation and in keeping with government specifications (U.S. Census, 1974), the elderly population is defined as those individuals who have attained 65 or more years of age. Among the disciplines recognized as having limited knowledge on behavioral attributes of the elderly was audi- ology and speech sciences. Consequently, the American Speech and Hearing Association delegated a committee whose primary objective was to conjoin with government officials to develop a conceptual framework which assessed the communication dis- orders existent among the elderly and to provide intervention strategies that took into account the special needs of geri- atric populations. The strategy for assessment and inter— vention was outlined in the following statement: Further investigation is required in the identification, evaluation and treatment of communicative disorders of the aged. Multidisciplined studies of medical, social and psychological aspects of aging are necessary for specifications of the meaning of rehabilitation in geriatrics, and for the development of a philosophy regarding communicative disorders of the aged. (National Institute of Neurological Disease and Stroke, 1968, p. 7). Past research investigating the speech and language of elderly persons has focused on how neurological diseases affect the communicative process. The disease that has cmmed extensive linguistic disability among the aged is stroke. Synonymously termed cerebral vascular accident, a stroke results from disruption of the blood supply to the brain due to cerebral thrombosis, hemmorrhage, or embolism (Felton, Perkins and Lewin, 1966). An epidemiological survey by Stallones et al., (1972) revealed a positive correlation between the occurrence of stroke cases and increasing age. Based on a sample of one million people, Stallones reported an estimated incidence of Stroke per 1,000 persons and expected new cases per year (Table l). A comparison of the incidence figures observed in group D with those observed in groups E and F indicates that a person's chances for having a stroke doubled at age 65 and more than tripled by the age of 75 years. The number of new cases per year followed similar trends. In addition, Stallones stated that 90 out of every 100 persons surviving the initial cerebral vascular accident were affected by post- strOke disabilities which required special assistance. omH.H oo.om as one me a omm oo.m whimm m omm om.m vmlmm a VHH oo.H vmlmv 0 mm mm.o velmm m o 00.0 vmlo d ummw mom mommo ooo.H mom mmcmm was macaw 3oz pmuommxm mocmpfloaH pmmeepmm ismm .a .msmav .qa em .mmzoqqaem em omemommm mm mZOmmmm ooo.H mmm mmomem no mmmmo zmz omaommxm ozm mozmoHozH omeazHemm H qumFNB The poststroke disability that is of primary concern to the speech and language pathologist is aphasia. According to Schuell and Jenkins (1961), aphasia is a reduction of language resulting from brain injury. This reduction cuts across all language modalities. Buckman (1971), Spahr (1971) and Sheridan (1976) recognized aphasia as having a high degree of prevalence among older persons. Because of its prevalence, aphasia is considered to be a geriatric communication problem. An abnormal linguistic behavior resulting from aphasia that has captured the attention of numerous investigators is semantic confusion. According to Schuell and Jenkins (1961), semantic confusions occur when an aphasic individual substi— tUtes a word that is associated in meaning or experience for the target word. Researchers investigating semantic confu— Sion among aphasic patients (e.g., Spinnler and Vignolo, 1966; Spreen, 1968; Rinnert and Whitaker, 1973; Spreen and Wachal, 1973; Zurif, et al., 1974) have found that 1) seman— tic confusions constitute the most frequently occurring error— tYpe among aphasic patients; and 2) the errors are not random, but appear to be substitutions of closely associated words for the target word intended by the patient. Responses given by aphasic subjects are usually compared to normative data cOmpiled via word association tests. In general, tests of Word association are categorized according to the type of response elicited from the subject (Cramer, 1968). In a free association test (Type 1) there are no circumscribed categories designated in which a response must be given. The subject is allowed to give a response from any semantic or conceptual category. In contrast, a restricted or controlled association test (Type 2) places a limit on the type of re— sponse that is acceptable. Information on the type of re— sponse required is either presented in the instructions (e.g., "You are to name the opposite of down.") or the sub— ject must select his answer from the response alternatives provided by the experimenter. Current terminology indica- tive of the two types of responses obtained from tests of word association has been proposed by Wiig and Semel (1976). Divergent semantic production (defined as the process of recalling and producing a variety of words and concepts, word associations, phrases or sentences) is synonymous with responses obtained on free word association tests. On the other hand, convergent semantic production (defined as the process of recalling or producing a specific word, word association, phrase or sentence to fit the meaning of the stimulus) closely resembles responses obtained on restricted association tests. A substantial amount of data has been compiled on free association responses of elderly persons. Dorken (1956) studied the oral responses of subjects whose ages ranged from 10 to 79 years. Data indicated a progressive increase in the number of common primary responses (i.e., the most frequently occurring responses given to each sti- mulus by each subject) until age 39. After age 39, however, the frequency of common responses began to decline. The 6 author attributed the resulting change in primary response strength to the subjects' active participation and involve- ment in the affairs of society until the late thirties. This was followed by a gradual withdrawal from certain activities accompanied by a decline in the occurrence of common associ- ative responses. Tresselt and Maynzer (1964) presented the 5223f Rosanoff Word Association Test (1910) to 738 subjects (ages 18 to 87 years). Results indicated a significant decrease in the pOpularity of the most common responses from the younger adults (ages 18 to 21 years) to the older adults (ages .55 to 87' years). In addition, more response variability (termed heterogenous responses by the authors) was found among the elderly group of subjects when compared to the younger group. Tresselt and Maynzer speculated that "the increase in variability (among the older adults) might be due to the development of rugged individualism in the older group as contrasted with a drive to conform in the younger group". (Tresselt and Mayzner, 1964, p. 66). In a study of 500 subjects in northern Germany, Riegel and Riegel (1964) verbally presented a lZO-item free Word association test to each individual subject. Groups 0f 76 subjects were equally divided by sex into five age levels (55-59, 60-64, 65-69, 70-74, and 75 years and over). Stimuli were divided into four grammatical classes and varied in frequency of occurrence in the German language. A compari- son of the older adult group with a young group (consisting of 120 subjects, ages 16-20 years, who were equally divided by sex) indicated a continuing increase in response varia- bility with age. In a subsequent study, Riegel and Birren (1965) found parallel results of greater response variability among the older subjects (ages 60 to 80 years) when compared to young adults (ages 18 to 33 years). Sefer and Henrikson (1966) administered a free asso— ciation test to 50 aphasic and 50 nonaphasic individuals, all of whom were in-patients at a veteran's hospital. Sixty- four percent of the nonaphasic subjects were over age 50. Comparing the nonaphasic subjects with an undergraduate male population used as subjects in a study by Deese (1962), the authors found a decrease in the number of homogenous responses given to certain form classes by the older nonaphasic adults. Sefer and Henrikson concluded that the difference in normal patterns of responses is affected by educational level, by frequency of stimulus words in general usage, and possibly by the subject's age. Comparatively few investigations of restricted asso- ciations have been conducted on nonpathological elderly per- sons. For purposes of the present investigation, a nonpath- ological elderly individual is operationally defined as a person, age 65 years and above, who passess the screening procedures of having: 1) no more than two errors on the Mental Status Questionnaire (Kahn et al., 1961), 2) normal hearing, 3) task-appropriate visual ability. 8 An elaboration of the screening procedures appears in Chapter Three (Research Methodology). The majority of published normative data on restricted association tests have been obtained by using college students as subjects (e.g., The Michigan Restricted Association Test, Riegel, 1965). Yet, tasks of restricted association have been presented to aphasic patients on several occasions in research (e.g., Spinnler and Vignolo, 1966; Lansdell, 1973; Lesser, 1973; Yamadorii and Albert, 1973; and Zurif, et al., 1974) and as subtests on formal diagnostic examinations of aphasia, (e.g., the animal naming subtest of the Boston Diagnostic Aphasia Examination, Goodglass and Kaplan, 1972). Although many of these studies do not report the age of their aphasic subjects, prevalence figures of aphasia among the elderly (Stallones, 1972) allows one to speculate that some of the patients fell within the age range of 65 years and above. Pragmatic application of word association research has evolved through a therapeutic procedure called "associa— tive cueing." During this intervention strategy, the clinician either presents associated words as cues (based on normative data) or capitalizes on the client's self-generated associa— tive responses. Berman and Peele (no date) hypothesized that self—generated semantic confusions often triggered the response intended by the aphasic patient. Reported case studies indi- cated that the self-generated responses consisted of synonyms, opposites and homonyms. Analogous categories have been used on restricted word association tasks (e.g., Schuell and 9 Jenkins, 1961; Riegel, 1965). In addition, Wyke (1962) provided experimental evidence which indicated that correct verbal responses from aphasic patients were facilitated when the possible choice of response words was restricted by the stimulus situation. A psycholinguistic theory proposed by Schuell and Jenkins (1961) stated that aphasia is characterized by a re— duction of language that follows the same principles as nor— mal language usage among adult populations. In support of this theory, Howes (1964) and Taylor (1966) found that the vocabulary of aphasic patients shifted in the direction of reduced variety. The reduction, however, followed the same type of word association patterns elicited from non aphasic Unormal") subjects. Results indicated that associative dis- turbances were a matter of reduced quantity rather than dis— similarity in the quality of responses given. Subsequent studies by Sefer and Henrikson (1968) and Rinnert and Whittaker H973) have reported similar findings and conclusions. The major criticism of these studies is that "normal language" is either not defined or is based on language samples obtained from young educated college students. Noting the accelerated increase in the number of persons attaining age 65 years, Rockstein (1975) stated that there is a need for a complete understanding of the facts of normal aging. In agreement with Rockstein, Hutchinson and Beasley (1976) Concluded that future investigations of language and speech functions in the elderly population should be aimed at 10 defining "normal" functioning. Such research could, perhaps, detect subtle syntactic, semantic and motor programming changes that result from the aging process. Alluding to the specialized instruction necessary for caregivers of poststroke victims, Oyer (1976) stated that knowledge regarding language behavior in the older population is limited and further re— search is required. Changes in word meaning from one generation to the next have also been considered in the aging process for language and speech functions. Recognizing word meaning as time and culture bound, Oyer (1976) speculated: If generational differences have attributed different meanings to the same words, ideas are not communicated any more than they would be were communicants speak— ing two foreign languages. Seman- tic obstacles can impede communi— cation between generations or between social classes or ethnic groups.(0yer and Oyer, 1976, p. 50). In summary, then, limited information is presently available on normal speech and language characteristics of elderly People. More specifically, results of restricted association tasks presented to elderly aphasic individuals are limited in scope because of the comparatively meager knowledge of restricted association responses given by nonpathological elderly persons. Considering the fact that educational level and possibly age affected the responses obtained from non— aphasic elderly subjects on a free association test (Sefer and Henrikson, 1968), there is a need to analyze the re- stricted associative responses of nonpathological elderly 11 persons within specific age levels and at varying levels of formal education. Purpose The purpose of the present investigation was to provide data on restricted word association tasks presented to a sample of nonpathological elderly persons. Before pre— senting the specific research questions to be investigated, an understanding of related terminology is necessary. Terminology For purposes of this study, the terms used are oper— ationally defined as follows: Lexical Stimulus Items. The 25 nouns selected from the original test of the Michigan Restricted Association Test (Riegel, 1965). The 25 nouns were presented to each of 132 subjects to elicit a set of responses. The list is included in Chapter Three (Table 5). Logical Tasks. This term implies that the response set given by each subject should have reflected class member— ship or class relationship to the lexical stimulus item (Flavell, 1963). The two logical tasks under investigation included superordinates and similars. Infralogical Tasks. This term implies that the response set given by each subject should have reflected a temporal or spatial relationship to the lexical stimulus items (Flavell, 1963). The two infralogical tasks under 12 investigation included locations and parts. Superordinate. The name of the general class to which each lexical stimulus item belongs (e.g., the class name for stove is appliance). Similar. A word that means the same as each lexical stimulus item (e.g., a word that is similar to car is automobile). Location. The place where each lexical stimulus item can be found (e.g., a lion can be found in the jungle). EEEE‘ An attribute or characteristic of each lexical stimulus item (e.g., part of a zebra are its stripes). Primary Response. The most frequently occurring single response given by the subjects to each test stimulus item. Common Response. A response given by two or more subjects to a test stimulus item. Idiosyncratic Response. A response that is given by Only one subject. Diversity Score. The number of different responses given by subjects to a test stimulus item under a specific task instruction. Research Questions Based on the above terminology the following research questions were investigated: 13 1. In the restricted association task of superordi- nation, what are the frequencies of occurrence of the common responses given by 85 percent of the nonpathological elderly group sampled? 2. In the restricted association task of similarity, what are the frequencies of occurrence of the common responses given by 85 percent of the nonpathological elderly group sampled? 3. In the restricted association task of location, what are the frequencies of occurrence of the common responses given by 85 percent of the nonpathological elderly group sampled? 4. In the restricted association task of part, what are the frequencies of occurrence of the common responses given by 85 percent of the nonpathological elderly group sampled? 5. In the restricted association tasks of superordi— nation, similarity, location and part what primary responses occur for each of the 25 lexical stimulus items? 6. Is there a significant difference between sub- jects, ages 65 to 74 years, and subjects, ages 75 years and above, on restricted association tasks of superordination similarity, location and part as measured by the number of primary responses given by each subject in each age group? 7. Is there a significant difference between male and female subjects on restricted association tasks of superordination, similarity, location and part as measured 14 by the number of primary responses given by each subject in each sex group? 8. Is there a significant difference between the subjects having less than high school education and subjects having education equal to or greater than high school on restricted association tasks of superordination similarity, location and part as measured by the number of primary re— sponses given by each subject in each education group? 9. Does the pattern of restricted association responses differ significantly between the logical instruc- tion Superordination, similarity)and infralogical instructions (location, part) as a function of age, sex or education? Organization of the Dissertation Chapter One contains a statement of introduction about the limited information available on the linguistic behaviors of nonpathological elderly persons. The exact purpose of the study and its potential importance to the research literature are included. The major terms used in this study are defined, and the research questions are stated. Chapter Two reviews the scientific literature per- tinent to this investigation. Theoretical and research papers investigating the cognitive ability of elderly indi- viduals are summarized. Of particular interest is the grow- ing concept of life-span developmental psychology. Previous studies of word association investigating the effects of age, 15 sex, educational level and stimulus presentation on responses elicited are reviewed. Chapter Three describes in detail the subjects, the screening tests, the equipment, the materials and the research procedures employed in this study. Chapter Four presents the data collected from this investigation and a discussion of the results. Chapter Five incorporates a summary of the problem investigated, the conclusions drawn as a result of this study and the recommendations for further research activity. CHAPTER II REVIEW OF RELATED LITERATURE Introduction Prior to World War II, the concept of aging in the American society was primitive in scope. Unlike persons living in third world countries of Africa and Asia, who have invariably viewed the aging process as an interaction among equally weighted physiological, psychological and sociocul- tural forces (Manney, 1975), the perspective of later life that was widespread throughout the United States was one of an irreversible biological process that caused systematic deterioration in the mental and physical abilities of an individual (Baltes and Willis, 1975; Botwinick, 1973; Kimmel, 1974). The post World War II introduction of life span developmental psychology, however, led to a revamping of the theoretical viewpoint which purported that decrement and deterioration were natural consequences of growing old. Essentially, supporters of life span developmental theory b-g., Baltes, 1973; Baltes and Willis, 1975; Busse and Pfeiffer, 1969; Goulet and Baltes, 1970; Kimmel, 1974; Labouvie—Vief, 1976; Schaie and Gribbin, 1975) contended 16 17 that both maintenance and decline of human abilities across the life span were a function of the interaction of physio- logical, psychological and sociocultural factors. The per— spective purported by life span developmental theorists was summarized in the following statement: The conclusion that aging change is change toward slowness, less behavior, less acquisition, less performance and greater dependency may be more a func— tion of the theoretical orientation applied than a representative assessment of the universe of aging change itself. (Baltes and Willis, 1975, p. 14). The authors suggested that both biological and environmental influences operate and interact in the production of behavior change processes. They concluded that greater empirical research efforts were desirable in strengthening environmen- tal, psychobiological and ecological conceptions of behavior change processes. The intent of the present literature review is to focus on theories and empirical research which have investi— gated cognitive functioning among elderly persons and to relate these findings to a particular component of language usage by the elderly. Specifically, the topics to be dis- cussed include the following: 1. Historical and Current Conceptualizations of Adult Cognition 2. The Relationship of Semantics to Word Association 3. Effects of Age, Sex and Educational Level on Responses Obtained on Restricted Word Association Tasks 18 4. Effects of Mode of Stimulus Presentation on Word Association Responses Historical and Current Conceptualizations of Adult Cognition Cognition is defined by Arenberg (1968) as effect- iveness in dealing with information. Processes of cognition include registering, storing, retrieving and manipulating information to solve a problem. Formal measures of cognitive abilities within individuals are obtained through the use of intelligence tests. Numerous studies in experimental psychology have focused on changes in cognitive skills across the life span. In fact, no area of aging research has received greater attention than the assessment of cognitive abilities via intelligence tests (Botwinick, 1967; Eisdorfer, 1969; Riegle, 1973a). Conclusions evidenced in most published research on adult cognitive skills are based on the bio— maturational ontogenetic model of cognitive development. According to this model, adult cognitive development is an extension of early childhood development. It is assumed that during childhood and adolescence, an individual acquires stable psychological traits, abilities and competencies (Riegel, 1975). Development is indexed by pre—programmed biomaturational stages which follow a unilinear irreversible Proqression (Reese, 1962). Furthermore, an individual is said to reach biological maturity during the adolescent years With mental, physical and physiological capacities Operating 19 at a level of peak performance. Labouvie—Vief (1977) points out that research investigating the period of life that re— mained after peak performance had been reached (the period from young adulthood through old age) addressed two major questions: 1) Once adolescence is reached, how long is peak performance maintained? and 2) When does performance become disrupted by presumptive biological de— terioration? Both cross-sectional studies (one observation on multiple cohorts) and longitudinal studies (multiple obser— vations on one cohort) have attempted to answer these two questions. Within the realms of the two research strategies, quantitative and qualitative characteristics of adult intelli- gence have been reported. Depending upon the research design employed (cross-sectional or longitudinal), a difference exists in the results obtained. In a classic cross-sectional study of adult learning, Thorndike (1928) determined that peak learning performance occurred at age 22 years and declined about one percent per Year until age 40. Similar results were reported by weschler (1958). During an attempt to standardize the Bellevue Intelligence Scale (which was later renamed the WEschler Adult Intelligence Scale), Weschler reported a high peak of performance for his subjects at 22 years followed by a gradual decline. Subsequent cross-sectional studies (e.g., Balinsky, 1941; Lienert and Crott, 1964; Schaie, 1962) 20 confirmed Thorndike's and Weschler's findings and extended the decline into the sixth and seventh decades of life. Investigations employing longitudinal research strategies have failed to substantiate the conventional view of decline in intelligence during adulthood and old age. Data indicated that increments existed in some subjects until at least the mid-fifties (e.g., Owens, 1953; Bayley and Oden, 1955). Other longitudinal research (e.g., Miles, 1934; Jarvik, et al., 1962; Kallman and Jarvik, 1959; Tuddenham, et al., 1968) supported the trends found in cross-sectional studies but concluded that the rate and magnitude of decline was less than was indicated by cross—sectional data. Thus, the conclusion purported by most investigators vms that cognitive performance among the elderly was charac- terized by irreversible regression and decrement. Further- more, the decline in cognitive function was attributable to rmrmal universal biological breakdown (Labouvie-Vief, 1976). The study of differential patterns of cognitive changes associated with adulthood and old age provided quali- tative data on specific areas of maintenance and decrement in intellectual capacities. Research centered around a nmdel conceptualized by Cattell and Horn (Cattell, 1963; 1357; Horn, 1968; 1970). Two differential patterns of cog- rfitive functions were distinguished: fluid intelligence and crystallized intelligence. Fluid intelligence referred to abilities which reflected neurological and physiological capacities and incidental learning (Hooper, Fitzgerald and Papalia, 1971). According to Cattell (1968), cumulative 21 effects of aging as well as environmental insults, loss of sensation in the sensory modalities, and insults to the brain were reflected in decreased fluid intellectual capacities. In comparison, crystallized intelligence was recog- nized as the product of acculturation and experience. This form of intelligence included individual differences in factors such as personal adjustment and motivation (Hooper, Fitzgerald and Papalia, 1971). According to Hooper and his colleagues, crystallized abilities were superimposed upon the basic (fluid) intellectual capacities, which, according to Horn (1968), operate to some extent in all intellectual performances. Cattell considered crystallized intelligence to be "the collection of skilled judgments a person has ac— quired by applying his fluid intelligence to his school Opportunities" (Cattell, 1965, p. 304). According to Labouvie-Vief (1976), the fluid- Crystallized model of intellectual ability has been sub- stantiated across literally "hundreds" of research studies. Emery major cross—sectional study conducted has consistently demonstrated that tests measuring stored information exhibit the least differences between young and old subjects. Con- verseLy, the earliest and most dramatic losses have been (xmsistently shown to occur on tests of memory, space and abstract reasoning abilities (Baltes and Labouvie, 1973; Botwinick, 1973) - Labouvie-Vief (1976) summarizes the distinct polarity between maintenance and improvement of crystallized intelligence 22 and deterioration of fluid intelligence in adulthood and old age. Speaking of fluid intelligence, she states: ...tests relating to the perception of abstract relationships among symbols of low meaningfulness (such as, for instance, geometric shapes) to the integration of new and complex materials and to the effective use of information under conditions of time restrictions and in highly abstract contexts are those on which the older adults tend to do much poorer than their younger counterparts. (Labouvie-Vief, 1976, p. 7). In.comparison, crystallized intelligence is viewed in the following manner: ...cognitive tests which assess an individual's accumulation of verbal skills and general information on tasks of learning, memory and prob- lem solving are those typically found to improve throughout adult- hood and well into old age. (Labouvie—Vief, ibid). Thus, as shown by Botwinick (1967), the classic pattern of cognitive ability among the elderly on, for example, the Vkmchler.Adult Intelligence Scale (WAIS) shows high scores r purposes of clarity, both fluid and crystallized intelli- gfince were discussed, the present investigation focused ex— clusively on an aspect of crystallized intelligence — the SEimantic relationships that exist for a given set of words. Thus, in section two of the literature review, a blTief discussion of the concept of semantics is provided, ltlle relationship of semantics to tests of word association j~53 discussed, and research focusing on changes in word asso— czjuation responses as a function of age, sex and educational leVel is reviewed . 28 The Relationship of Semantics to Word Association Theories of modern linguistics conceptualize language as a form of structuralism (Fillebaum and Rapoport, 1971). Linguistic structuralism implies that each language is re- garded as a set of interrelated systems characterized by linguistic elements (e.g., phonemes, morphemes, grammar) and the relations that exist among these elements (Lyons, 1968). The relationships between phonological and grammatical systems Of the English language have been formally documented by such linguists as Chomsky. In his book, Syntactic Structures, Chomsky (1957) substantiates the notion that phonological and grammatical structures of a language can be studied objectively because of the finite and concise rule system implicit to these two linguistic elements. In contrast to the objectivity and finite character of phonology and grammar, the semantic element of a language is composed of subjective lexicons (Fillenbaum and Rapoport, 1971). According to these two psycholinguists, a lexicon (Or word) is considered to be subjective because its meaning is intricately woven into the perceptions and understanding of the external world by human observers. Subjective lexi— QOns are not isolated independent features of a language. Irlstead, the words are combined to form interrelated associ— a“tive structures which constitute the semantic domain of a la-I'iguage and represent the "world picture presupposed by those who speak that language" (Miller, 1968, p. 68) . 29 Associative networks, defined as a relationship among more than two elements, are composed of lexical fields con— sisting of a given lexicon and other words which are closely or remotely related to the lexicon conceptually (Miller and Johnson—Laird, 1976) . According to Miller and Johnson—Laird, a lexical field is composed of two processes. First is the process of shared conditions that determine denotations of the words within the field. Second, there exists a concep— tual core which is an organized representation of general knowledge and beliefs about what objects or events the word denotes. The internalized configuration of the relationships Comprising the sematic domain has been of interest to experi— mental psychologists for many years. Empirical investigations of the semantic domain have mainly been conducted through the use of word association tests. Introductory comments about word association was included in Chapter One. Thus, the remaining portion of the literature review will summarize Studies that investigate changes in word association responses as a function of the subject's age, sex and educational level. Studies which investigate the effects of sensory modalities L‘lSed for presenting the stimulus items on the responses ob- tained are also reviewed. Lastly, the procedures employed by investigators to analyze data obtained from tests of word aSsociation are discussed. 30 Effects of Age on Responses Obtained on WOrd Association Tasks The increasing popularity of life span developmental psychology has motivated researchers to explore associative networks in middle and late life. Extending Piagetian theory irrto adulthood, Riegel (1973 b & c) proposed that the seman— tix: domain in adulthood and old age is controlled by "dialec- tic: operations". As defined by Riegel, dialectic operations rxep>resented simultaneous usage of the skills existing at all fcnar of the cognitive stages originally proposed by Piaget. Thuese stages include the sensori—motor, preoperational, con— CIHete operational and formal operational stages of cognitive d-€B'V'elopment. Piagetian theory (Flavell, 1970) states that arr individual who experiences disequilibrium strives toward equilibrium by utilizing formal operational thought processes aJId, thereby, moves away from contradictions and concrete tllought processes. In contrast, Riegel's theory of dialectic C’IPerations proposes that a mature individual accepts disequili- blfium and tolerates contradiction. Therefore, the individual ‘Céin transcend different levels of cognition simultaneously. Eiiegel purports that such "multilevel synchronicity is desi- 1Table for appropriate performance in different situations" (Iiiegel, 19730, p. 482). Previously dormant as a consequence (Di? physical restrictions, normative social pressure and formal GECiucation, this newly acquired flexibility in cognitive skills $3l'lhances creative thought. According to Riegel the creativity E13nd flexibility of thought processes in later life perhaps a~<2counts for the extensive response variability on tasks of 31 word association reported in the literature. Studies of free associative responses characteristic of nonpathological elderly subjects have been summarized in Chapter One. The review which follows discusses results ob— tained on restricted association tasks that were administered to older subjects. In a reverse procedure, Riegel and Riegel (1964) asked 12 subjects (age 64 and over) to identify the original stimuli (called target words) that had produced the responses obtained On the Michigan Restricted Association Norms (Riegel, 1965) . The experimenters used responses collected on the norms as Stimuli (called clue words). The eldery subjects' responses W‘Ere compared with 96 young subjects. For both young and old Stflojects, double clues were found to produce more correct idEntification of the target word than single clues. "Sjnnilars" were found to be the most efficient single clue. Ir1 addition, simultaneous usage of similars with any of the cDther clues (location, superordinates, and parts) was found ‘tC> increase correct identification of the target word. In (ZCanarison, locations and superordinates were very poor as Ssingle clues but were only slightly below the average score wklen used in combination with other clue words. Parts gave a‘7erage correct responses regardless of their being used as 6i Esingle clue or in a combination with another word. Grouping the clues into logical (superordinate and Ssil‘nilars) and.infralogical (locations and parts) components, II‘iegel found that double logical clues showed large age 32 deficits, whereas both double and single infralogical clues produced less marked declines. These findings are in direct opposition to the results obtained by Zivian (1966) and Wigg and Globus (1971). Both studies reported that a greater runnber of target words were correctly identified from logi- caLL~clues than from infralogical clues. Mixed clues (e.g., pa1:ts and similars) showed average decline. Riegel (1964) cxaricluded that the findings implied a specific lack of <2aIpacity on the part of elderly subjects to deal with more akJStract (logical) relations. The elderly were said to pre- fear'a concrete (infralogical) focus of conceptualization. Wigg and Globus (1971) performed a similar associ- ailion task using 11 aphasic subjects and 11 college students. que two groups were matched for sex and educational level. Ages were not reported. Twenty nouns, selected for their picturability, were clnosen from the list of the original stimuli used in the Michigan Restricted Association Norms. The clue words were CElassified into two categories: logical clues and infralo- giical clues. Logical clues were selected from the super— c>3:‘dinate, subordinate and similar response categories. Infra- lOgical clues were selected from the location, parts and pre- C=62ding response categories. No overall quantitative difference er1 responses was found. However, qualitative analysis of ElE>hasic and non—aphasic performance suggested that facilita- tli.on of target word identification was similar. For both SSllbject groups, the largest number of correct target words VVEis given in response to high logical clues (high implying 33 associated strength based on tabulated frequency of response occurrence). Both subject groups identified the smallest number of target words when given low logical clues as stimu- li. Although associative strength of logical clues was not neported by Riegel (1964), the discrepancy in the results okrtained from these two studies may be a result of associa- tixre strength being confounded with the logical clues selected. Effects of Sex on Responses Obtained on Word Association Tasks Sex differences on responses to word association tEISks have been given limited attention. The subjects who ENirticipate in studies are usually equally divided by sex, kNJt data are reported with no consideration of sex as an in— dependent variable . A few studies of free word association investigated SEEK differences in the 1960's. Palermo (1963) and Palermo 311d Jenkins (1965), investigating several age levels from gutade one through college, found that females have a greater number of primary responses than males. A primary response iJS the most frequently occurring response to a given stimulus. UDlle authors also found that a significantly greater percentage C>I5 females gave the first three most popular responses in t:}1e sequential order found in previously collected response 1Tlierarchies. Additional findings in studies by Palermo (3—963) and Palermo and Jenkins (1965) showed that males give 34 more supraordinate (class) responses than females, irrespec— tive of grade level. Effects of Educational Level on Responses Obtained on Word Association Tasks Relatively few studies have investigated response vairiability as a function of formal educational level. The diifficulty with investigating this variable is that age and edilcational level are confounded. However, in a related Stlldy of classification using shape, size and color, Denny (1J974) suggested that the effects of the highest level of eéhlcation reached on classification abilities should be iJIVestigated. Herr (1957) gave 1600 adults an association test wi.th instructions which asked the subjects to give the re- SENDnse most people would give. The investigator found a Snuall but significant positive correlation between response cCHumonality and educational level. Becher (1960) found greater response commonality an[long female college freshmen than among college seniors. ““1631 the freshmen were tested four years later, the females VVEire found to have a lesser number of common responses. Comparing 50 adult aphasic patients with 50 subjects VVIIC) were admitted to a hospital for reasons other than neuro- l<><3ical disorders, Sefer and Henrikson (1966) stated that (ijLifferences in normal patterns of word association responses WEEIre, perhaps, affected by the educational level of the sub— :JEE<=ts as well as several other factors. 35 Rosenzweig (1964) investigated the primary responses of 115 French workmen. He speculated that perhaps the edu— cational and socioeconomic levels of the workmen resulted in systematically different associative responses. Previously collected data on college French students were used for purposes of comparison . Effects of Mode of Stimulus Presentation on Word Association Responses Two modes of stimulus presentation, aural and visual, hanre been used extensively by experimenters investigating WCHfd association. Aural presentation is generally conducted bY’ the experimenter's pronouncing the stimulus item aloud. ViSually presented stimuli are either printed on a page or pIKDjected on a screen (Cramer, 1968). Two dependent variables have been investigated in Stflldies that exclusively looked at effects of mode of stimu- 1113 presentation: 1) the number of responses obtained and 2) the type of responses obtained when input modes varied. Reynolds et al., (1971) compared differences in Esingle word responses of orally—~ versus visually-presented Silinmfli at four different age levels. Forty subjects were raIldomly drawn from grades 3, 5 and 7 and an adult group 1:ieil‘lging in age from 21 to 92 years. Groups were balanced :EC>IT sex. All subjects were described as educationally subjects were presented ci€3E>rived. During oral presentation, t:}1€e stimuli aloud by the experimenter. Stimulus items were E31Teesented visually by using an overhead projector with the 36 words printed on transparencies. The third grade children were found to have more single word responses when the oral mode of presentation was used. For each successively older group, response differences between oral and visual presen- tation of the stimuli were negligible. Ohnmacht and Pacheo (1972) administered a word association test to 114 literate adults to determine the effects of mode of presentation (aural vs. visual). Half of the subjects were presented 12 concepts aurally; the remaining half were given the same 12 concepts visually. Results showed that the quality of associative response was unaffected by the mode of presentation for literate adults. Subjects were found to respond with a single word regardless of modality. A significant difference in the number of associations generated by the subjects was found to be in favor of the aurally presented stimuli. However, the authors speculated that the specific stimulus items used may have been a confounding variable. Timnermaus and Kumin (1974) used 10 adult male aphasics and 10 normal males, matched for age and education, t0 investigate free association responses of adult aphasics tO auditory and pictorial stimuli. The stimulus items con— Sisted Of 40 nouns and verbs from the Palermo—Jenkins list (1964). All stimulus words had a high frequency of occur— rence. Puohasics and normals were found to emit significantly Imore SingJJe word responses on auditory than on visual modes of Stimullls presentation. Normal subjects gave significantly 37 more single word responses than the aphasic subjects and more syntagmatic responses for the visual and the combined mode of stimulus presentation. The authors concluded that perhaps the auditory association and visual association tasks were not comparable. According to Timnermaus and Kumin (1974), when the visual association modality was used, the subjects appeared to label the stimuli orally first and then to give an associated response. In contrast, when stimuli were presented aurally, the subjects rarely imitated the word before emitting an associative response. The effects of mode of presentation on responses obtained from elderly subjects have not been investigated. However, McClusky (1970) states that the most effective learning environment for the elderly appears to be one which combines auditory and visual information. Procedures Employed to Analyze Data From Word Association Tests Associative responses which are elicited by a word aSsociation test must eventually be categorized in some manner. A large variety of measures have been used to describe the findings of word association studies. However, Only those analytical procedures employed in the present inveStigation will be discussed. The measure that is used most often in analyzing word aSSOCiation data is the primary response (Cramer, 1968). The PriHEiry response is the most frequently occurring response given bY’ a group of subjects to any one stimulus. The 38 criteria for what constitutes a primary response may be based either on previously gathered normative data or on the sample currently being tested. In addition to analysis of word association data via primary responses, other quantitative measures also reflect upon the organization of the associative network. In general, these measures indicate the size of the associative network. One scoring procedure that is used to measure the size of the network is the number of different responses obtained across subjects on a given stimulus item. The score obtained is known as the diversity score and is symbolized by the letter "D". In contrast to primary response analysis, which determines response commonality or homogeneity, the diversity score indicates response variability or heterogeneity. Response entropy (calculated by the formula H = — pi hagzpi where pi is the probability of occurrence of a given response) has also been used to describe the heterogeneity of responses to a stimulus word (Cramer, 1968). However, Laffal (1955) demonstrated that response entropy and diver— sity measures were highly correlated. Thus, the diversity score is the more preferable measure for determining the size 0f the: associative network because it is easier to calculate than Iwesponse entropy and is as reliable as the latter measure in calxzulating the variability of responses in an associative domain (Laffal, 1955). 39 Summary The second section of the literature review covered information on the concept of semantics and its relation to tests of word association. Studies investigating changes in the type of responses given on word association tests when subjects were grouped by age, sex and educational level were reported. Investigations which reported on how experimenter manipulation of the sensory modality used to present the stimuli to subjects (usually aurally, visually or a combina- tion of aural and visual presentation) affected the responses elicited, were summarized. Because of the limited number of investigations which concentrated on these four variables (age, sex, education, mode of stimulus presentation), evi- dence is somewhat inconclusive at this time. Finally, analytical procedures pertinent to the preasent investigation were discussed. The experimental procedures employed in the present study are reported in Chapter Three. _..—________ wiw CHAPTER III RESEARCH METHODOLOGY Subjects One hundred thirty-two elderly adults served as subjects in the present investigation. All subjects were white, native born American English speakers. The age range of these 132 persons (61 men and 71 women) was 65 to 89 years. Subjects were divided into two age groups: Young-Old (Y—O) (xmnposed of persons whose ages ranged from 65 to 74 years 6? = 70.21 years, standard deviation = 3.81 years) and (”xi—Old (O-O) composed of individuals aged 75 and above G? = 79.84 years, standard deviation = 7.10 years). Place- meqrt of the elderly subjects into two separate age groups was based on the classficiation system proposed by Busse and Pfeiffer (1969), who assigned the terms "young-old" and "old-old", respectively, to the age ranges indicated. Aldunigh Neugarten (1975) expanded the age range of the YOung-c>ld group from 55 years to 75 years, she supported the ages range designated by Busse and Pfeiffer as old-old. Subjects were noninstitutionalized community resi- dents who participated in a senior citizens' community Program or recreational facility. The majority of the 40 41 elderly subjects resided in two government supported housing projects in Inkster, Michigan. The remaining subjects (most of whom were retired state employees or retired university professors) lived in East Lansing, Michigan. Subjects were classified by educational level on the basis of two criteria: less than high school completion ((HS) and high school completion or above (3H8). Screening Procedures Used The selection of the elderly individuals who were used as subjects in the present investigation was based on three screening procedures administered and evaluated by the investigator prior to the subjects' involvement in any formal research tasks. All subjects were tested for l) the condition of mental functioning as operationally de1:ermined by the score obtained on the Mental Status Questionnaire (Kahn, et. al., 1960); 2) task appropriate visual ability; and 3) normal hearing with intensity levels adjusted for presbycusis. The Mental Status Questionnaire is a quantifiable index of mental functioning. The index consists of ten qUeStjans (Appendix A) which quickly and accurately assesses an huiividual's mental abilities on tasks of memory, orien— tatiOn, calculation and retrieval of current information. A quantLitative measure of cerebral functioning can be deter— IniHEd OI: the basis of the number of errors obtained. Kahn 42 and his associates (1960) developed a chart which indicated the number of errors associated with the degree of organic brain syndrome found in individuals (Table 2). TABLE 2 NUMBER OF ERRORS ASSOCIATED WITH THE DEGREE OF ORGANIC BRAIN DYSFUNCTIONS AS DETERMINED BY THE MENTAL STATUS QUESTIONNAIRE Degree of Organic Number of Errors Brain Dysfunction 0 _ 2 none or minimal 3 _ 8 moderate 9 _ 10 severe Wilson and Brass (1973) have found the Mental Status Queastionnaire to be a powerful single measure for detecting arui roughly quantifying intellectual impairment. Included hi their study are references that show the test to be a valid and reliable measure for use with geriatric popula- tions. Citing the advantages of the test, Wilson and Brass stated.that the Mental Status Questionnaire 1) provided an Objectzive basis for the uniformity of observation and eval- Hatiori by different examiners; 2) required little time fOr adminisstration, making it a desirable procedure for rapid Clinicatl screening and for research purposes; and 3) in- Cluded.:five questions that relate to personal information Eind can, therefore, be asked without the patient's knowing 43 s/he is being tested. Although the Mental Status Questionnaire is a use— ful instrument for examining mental change associated with chronic brain syndrome, it is limited in detecting other types of psychiatric disorders (e.g., depression, schizo— phrenia, paranoia). A second disadvantage is that the test measures the mental faculties at the time of testing only. Thus, the Mental Status Questionnaire may fail to show im— pairment in persons with fluctuating mental ability. How- ever, since the present investigation is concerned only with the mental status of the subject during the period of testing, a fluctuating mental ability pattern was not considered. Criterion for acceptance as a subject in the present investigation was designated by a score of zero to two errors or1 the Mental Status Questionnaire, indicating no or minimal celrebral dysfunction. To satisfy the criterion for normal hearing, each participant passed a bilateral pure tone audiometric screening test given at the octave frequencies lOOOHz, 2000Hz, and 4000Hz via Telephonics TDH 3945. earphones mounted in MX 4l/AR cushions. Criterion for passing the hearing test was béised on data cited in a study by Milne and Lauder (1975) , who calculated loss of hearing acuity as a function Of agillg. This phenomenon, termed presbycusis, is defined as the hearing loss that occurs with age unconfounded by Extranenaus factors such as disease and noise exposure (Corso, 1971)- Selection of the three test frequencies was based on 44 guidelines for identification audiometry approved by the American Speech and Hearing Association Legislative Council in 1974 (Asha, 1975). The hearing threshold level (HTL) of each subject was tested in a quiet room with acceptable ambient noise levels. The allowable ambient noise level for each test frequency and the actual ambient noise levels observed in the three test environments are shown in Table 3. TABLE 3 COMPARISON BETWEEN THE ALLOWABLE AMBIENT NOISE LEVELS FOR lOOOHz, 2000Hz AND 4000Hz AND THE ACTUAL AMBIENT NOISE LEVELS OBSERVED IN THE THREE TEST ENVIRONMENTS Allowable Ambient Frequency Noise Level lOOOHz 50 dB 2000 58 4000 76 Observed Ambient Te ' ' st Env1ronment Neise Level Board Room (Inkster) 48 dB Chapell (Inkster) 43 Chapel2 (E. Lansing) 47 A General Radio Sound Level Meter (Type 1551-C) was used to determine the relative ambient noise levels of the test enVirOIHHents. Dials were set on "Weight A“ and "Slow Sweep" 45 to obtain average intensity peaks across all frequencies. A calibrated Maico Portable Audiometer (Model MA-ZB) was used to conduct the pure tone screening. To determine whether the elderly subject had ade— quate visual ability to participate in the research tasks, each elderly individual was asked to read aloud the general instructions which provided procedural information on the four tasks (Appendix B). The instructions appeared on an 8 1/2" x 11" piece of typing paper. The orator typing element of an IBM Selectric II typewriter was used to write out the instructions. The orator element types boldface upper and lower case letters (5mm and 3mm in height, respectively). Criterion for passing the vision test was the ability to read the instructions accurately with or without corrective lenses when the paper was held eight to twelve inches away from the subject's face at a 45 degree angle. Stimuli Twenty-five nouns were selected from a core of 200 words (120 nouns, 40 verbs and 40 adjectives) which comprised the original stimuli used by Riegel (1965) on the MiChigan Restricted Association Norms. Nouns were eXclusiVely chosen as stimuli becasue they represented the largest number of stimuli observed in an individual form class in Riegel's study. Prior to selecting the zslunnls, it was first necessary to determine the form 46 class (part of speech) and the frequency of occurrence of the 200 stimuli in the English language. Riegel reported cumulative data on the number of nouns, verbs and adjectives included in the list of the original stimuli. The number of words that appeared at each frequency of occurrence was also reported. However, Riegel did not specify the words that were assigned to the frequency levels or to the form classes. Adapting the same procedure used by Riegel, the form class of each of the 200 stimuli was determined by using the first part—of—speech entry listed in the Random House Dictionary of the English Language, Unabridged Edition (1973). The initial part-of—speech entry represents the most common usage of the word in English sentence constructs. A compar— ison of the number of stimuli listed in each form class by Riegel versus the figures obtained by the investigator showed a slight discrepancy in the number of words assigned to each fornlclass. Table 4 shows that there was a slightly higher number of words classified as nouns when the first entry was based on the Random House Unabridged Dictionary (1973). Perhaps the discrepancy may be accounted for by the fact that Riegel used Webster's Collegiate Dictionary to deter— mine the first entry for each word. The particular words that Shifted form class cannot be determined since the inforInation about individual word classifications was not included in the final report of Riegel's study. 47 TABLE 4 DIFFERENCE IN THE NUMBER OF WORDS CLASSIFIED UNDER THREE FORM CLASSES BASED ON THE FIRST PART“ OF-SPEECH ENTRY OF TWO AMERICAN ENGLISH DICTIONARIES Unabridged Form Class Webster Collegiate Random House (Riegel, 1965) (1973) noun 120 125 verb 40 39 adjective 40 36 The selection of the stimuli from the 125 words classified as nouns by the Random House Unabridged Dictionary was based on the following rules: 1. Independence— the 25 stimulus items could only minimally elicit each other as responses to stimulus presen- tations. To determine independence, the three most common responses of each noun in each of the four tasks under present investigation (superordinate, location, similar and Part) were analyzed using the original responses from the Michigan Restricted Association Norms (Riegel, 1965). 2. Concept Representation- the 25 stimulus nouns were to represent a variety of concepts. Included in the 25 nouns were common objects, bodily functions, professions, family Itelationships, bodily structures, foods and animals. Based Orl the two rules, the investigator selected 25 words as leXiIIal stimulus items. The 25 stimuli appear in Table 5. 48 TABLE 5 LIST OF THE TWENTY-FIVE TEST STIMULUS ITEMS SELECTED FROM THE ORIGINAL TWO HUNDRED TEST STIMULI OF THE MICHIGAN RESTRICTED ASSOCIATION NORMS (RIEGEL, 1965) anger bread cabbage candle city eagle girl house knife memory moth music nurse rug soldier stomach street sun table thief tiger tobacco water whiskey wool 49 Generation of Task Stimuli Each of the 25 lexical stimulus items was centered on a 9" x 12" sheet of paper using Chartpak Velvet Touch lowercase lettering (Style 60 PT/Mll6OL). A sample of the lettering appears in Appendix C. The large size letters were used to assist the elderly individual in visual recog— nition of the stimuli. A photograph of each stimulus item was taken by the investigator under the direction of a professional photo— grapher. The film was processed into 35 millimeter slides (black figure, white background). Four slides per noun were developed. Although the 25 lexical stimulus items were the same on each of the four tasks, ordinal presentation of the stimuli was arranged from the "easiest" to the "least easiest". Determination of the easiest to least easiest ordinal listing was based on the frequency of occurrence of the most common response of each of the 25 test stimulus items obtained from the 100 college educated subjects in Riegel's study. In the event that two or more words received the same frequency of common responses, the items were listed alphabetically. Thus, the number of presentations of the 25 teSt Stimulus items totaled one hundred. A master recording of the 25 stimuli was made using a White Hale talker (fo = 125 Hz) who spoke with a General Americarl Dialect. The talker was seated in a sound treated recordilig booth. An Electrovoice Dynamic Omnidirectional 50 Microphone (Model 635 A, frequency response = 60Hz — 15,000Hz) was placed eight inches from the talker on an insulated tabletop. An Ampex full track tape recorder (Model AG 600), located in an adjacent sound treated booth, was connected to the microphone. The recording intensity level of the talker was monitored by the volume control and a VU meter. Each test stimulus item was spoken thrice (using natural intonation and stress) at an average inten— sity level of —2dB VU to avoid distortion and peak clipping of the spoken signal. If, for any given item, the three productions did not satisfy the investigator, the item was repeated following completion of the entire stimuli list. The selection of the highest quality recording of earflistimulus item was performed by dubbing from the origi- nsil tape recording of the 25 nouns onto a second Ampex AL; 600 tape recorder. Using the VU meter of the tape reacorder and TDH 394‘. earphones, the three productions Off each stimulus were monitored both visually and aurally b)? the investigator and a student of speech science. An ir111ensity level of average peaks was the primary concern WTlean choosing the recorded version to be used as the test Sililnuli. Arbitrary volume levels were set to provide a Coristant intensity level (-2dB VU meter) of the stimulus Sixgnals. If all three recordings peaked at the appropriate intensity level, the decision for selection was made on the basis of interjudge agreement by the investigator and the SPeech science student on the best sounding production. 51 Following location of the best production, the sequence was . dubbed onto the Ampex AG 600 tape recorder four times. A unit designed for editing the dubbed audio tape was used to cut and splice seven seconds of silent interval space between each of the test items. Each second was re— presented by seven and a half inches of blank lead tape. Instrumentation for editing included a playback head con- nected to a preamplifier (Claricon, 36-195) routed to a speaker (Bogen, Model MTA-lO). IWith this unit, the beginning and end of each test stimulus could be located. The first and last second of each silent interval was marked with the name of the stimulus and the letters A, B, C or D (repre— senting the four tasks under investigation). Based on the ordinal listing of easiest to least easiest stimuli in each of the four tasks, the tape pieces were spliced together. A copy of the tape with the inserted silent intervals was pro- duced using an Ampex AG-600 tape recorder. The final master copy of the stimuli (presented to tlhe elderly subjects) was produced on two 15 minute Memorex cEissette tapes using the Montage Audio—Mate 600 cassette talpe recorder (frequency response = 80Hz to 10,000Hz). The Orkiinal listings for the superordinate and location tasks Wefire recorded on one cassette tape and the ordinal listings fCXr the similar and part tasks were recorded on the second Ca ssette tape . The Montage Audio—Mate recorder has two built in ENJlse signals: a lOOOHz pulse signal called the projector Command and a 3000Hz pulse signal known as the tape-stop 52 command. Synchronization of an audio signal (cassette tape recordings of the 25 stimuli) with a visual presentation (slides of the 25 test stimuli) is accomplished by record— ing the inaudible advance tone signal (lOOOHz) and the tape- stop signal (3000Hz) on the lower half of the tape (track two). The lOOOHz signal advances a slide projector, whereas the 3000Hz tape—stop command signal halts tape and projector motion until it is re—initiated by a remote control button. The position of the tape—stop command is based on a pre— selected point in the recording. Using a stop watch for temporal monitoring, the two inaudible pulse cues were programmed onto the cassette tape inbetween the stimulus items. Figure 1 indicates the sequence in which the silent pulse signals were inserted. Test Environments Three testing sites were chosen to conduct the investigation. The first two sites, located in Inkster, Blichigan, consisted of the board of directors meeting room (23'L x 11 l/2'W) and the L - shaped chapel (l6'L x 11 1/2'w X 28'W) of two government sponsored housing projects for the eldderly. The third site, located in East Lansing, Michigan, WEIS a chapel (l7'L x l7'W) in a local university affiliated ch‘urch. The selection of the test environments was based on 1) permission of the administrators of the facility to use tile site for testing, 2) convenience of the site for a sub— Stuantial number of elderly persons (e.g., free parking, easy .mEmUH msasafium “woe on» Mo GOHomucmmmHm HMSmH>\0HUs¢ pmufidonsosmm m msHCHmuno How mamcmflm UGMEEOU moum mama map can pcmfisou Houomnonm map mo soapummcH ecu mo cowomupmsoama .H musmflm .Houomm loam mpflam can Hmpuoomu mmmu map mum>fiuommu o» powmQSm way an pmmmmsm was omsu coofl3m muofimm mpofimm .wocmsvmm pcoomm cm>mm mum>m mo pcoo 1mm puflcu on» so pmuummca mm3 mmasm .Houoww loam ocean cam Hmpuoown menu on“ Umum>fluomcfi £0H£3 Hmcmwm mmaom pcmfifioo moum mmmp um ooom um xm .memmmm mpfiamsq amen m cadence poanSm comm 0p Maamusm poucmmmnm mnm3 flHSEflpm .mmmu muummmmo may no one xommu so pmpuoomu flaseflum H .m was m .mocmsqmm Ucoomm cm>mm mum>m mo pcoomm Sawfly map so pmunmm IGH mm3 omacm .Houomflonm mpflam may wmocm>©m coacS Hmcmflm mmasm pcmEEoo Hepomnoum um coca 3 5 mm Ma "GOHDMGMmem mpofimm m N H b m m w m N H h m m w m N H Amocoommv mem were warm Nana N gonna fie em H View: p 54 accessibility, knowledge of the site's location by the elderly), and 3) site was relatively quiet and uninterrupted by other scheduled activities. The ambient noise levels of each test site were reported in Table 3. Equipment Below is a description of the equipment which was employed in the present investigation, along with the arrangement of the equipment in the test environments. Two carrousel slide trays, each having a load capacity of 80 slides, were used to load the 100 35 milli— meter slides. Slides for Task A (superordinate) and Task B (location) were placed in one tray, whereas slides for Task C (similar) and Task D (part) were loaded into the second tray. Two 15 minute high quality low noise cassette tapes (Memorex) contained the serially ordered audio portion of tile test stimuli. The presentation of the stimuli on tape chincided directly with the slide presentations. Depending On the task being presented, one of the cassette tapes was loauded onto an Audio Mate 600 Cassette Tape Recorder that leati to a high quality speaker. The volume and tone levels 0f tile cassette tape were held constant (five and nine, reSPeCtively) throughout the presentation of the 100 test Sthmilus items to each subject. An Ektographic Slide ProjeCtor (Model AF-2) was also connected to the cassette 55 tape recorder. Both the slide projector and the cassette tape recorder were advanced in synchrony via a remote con— trol switch, leading from the recorder. The subject was seated in a comfortable chair facing a rearview plexiglass screen (22.5cmL x 28.5cmW) and within easy reach of the remote control switch. The cassette tape recorder, slide projector, speaker and rearview plexiglass screen were placed on a long table (6'L x 2'W). To control for response reliabilty, a second casette tape recorder (Sony, Model TC 60A) and microphone were located on the table. The microphone, leading to the Sony cassette tape recorder, was placed directly in front of the subject ten inches away at a 45 degree angle. A schematic diagram of the physical arrangement of the equip- ment in the test environment and the position of the subject is depicted in Figure 2. Test Procedure Each subject was seen individually for approximately 40 minutes. Subjects were given a printed c0py of five pages Of instructions faced down. The subject was asked to pick up the first page (General Instructions) and read aloud the Contents of the page. Questions about general procedures Were answered by the investigator following the reading. The subject was then allowed to read the instructions for Task A (superordination) and three practice items for Task A were presented. After the practice session was 56 Figure 2. Arrangement of Subject and Equipment in the Test Environment. = subject; b = loudspeaker; C = remote control; = microphone; e = plexiglass rearview screen; = Sony cassette tape racorder; g = Audio Mate 500 cassette tape recorder; h = Ektographic carrousel slide projector H1049) _. :4. ._.___.__ ._ __._.a._vevvw—v-w 57 completed, the subject was allowed a brief period to ask questions of clarification. The format of presentation of the test stimuli was the same for the remaining three tasks (instructions, three practice items, questions of clarifica— tion, presentation of 25 test stimulus items). Once the subject appeared to have an understanding of the test procedures and the linguistic responses requested, the first 50 test items (Slide Tray One) were presented, followed by a three minute rest period. The rest period provided a moment of relaxation to the subject to minimize fatigue. The first slide tray was replaced by the second tray (containing Tasks C and D) during the rest period along with the corresponding cassette tape. At the end of the short break, the remaining practice items and the last 50 test stimuli were presented to the subject. The 100 oral responses of each subject were re- corded on an answer sheet (Appendix D) by the investigator. In addition, the subjects‘ responses were recorded onto one Side of a 60 minute Ampex cassette tape using the Sony caSsette tape recorder. Subjects were given 30 seconds to respond to each item. If no response was elicited during the alloted time Period, the subject was instructed to advance the tape and Slide to the next test item. After all 25 items were pre- sented, the items missed within the task were presented a second time by the investigator acting as the talker. If 58 the subject could not provide the investigator with an associative response after the second presentation of the test stimulus items, a “NR2" (indicating no response after two presentations) was recorded for the missed item. Summary Chapter Three has presented an overview of the research procedures employed in the present investigation. One hundred thirty—two nonpathological elderly persons (ages 65 to 89 years) served as subjects. All subjects were white, native born American English speakers. Each subject had passed three screening procedures which assessed the subject‘s hearing, visual and mental abilities. Subjects were seen individually for approximately 40 minutes. Three testing sites were chosen to conduct the inves- tigation. The first two sites were located in two government Sponsored housing projects for the elderly in Inkster, Michigan. The third site was located in East Lansing, Michigan, at a local university affiliated church. The three sites Were selected on the basis of convenience for the elderly SUbjects. Using a synchronized audio-visual procedure, 25 nouns Were presented four times under two logical task conditions (Superordination and similarity) and under two infralogical task conditions (location and part). Presentation of the 25 noun stimuli varied from task to task as a function of the "easiest“ to the "least easiest" stimulus item under each task condition. 59 Written instructions were given for each task. Subjects were asked to elicit a one-word response for each noun under each task condition. The subjects were given 30 seconds to respond. Stimulus items receiving no response during the alloted time interval were repeated a second time by the investigator. If no response was given after the second presentation, an "NR2" was recorded indicating no response after two stimulus presentations. Responses elicited from each subject were graphically recorded onto an answer sheet and auditorily recorded onto a cassette tape. Chapter Four will discuss the statistical analysis and results of the investigation. CHAPTER IV RESULTS AND DISCUSSION This chapter is divided into two main sections, results and discussion. The section on results includes the descriptive and inferential statistical analyses to which the data were subjected along with an explanantion of what these analyses revealed. Data are represented graphically where appropriate. Finally, a discussion of the results as they relate to other relevant research is provided. Results Descriptive Analyses The descriptive analyses employed in this investi— gation provided both qualitiative and quantitative interpre- tation of the data. The first descriptive analysis to which the raw data were subjected was Crosstabulation Analysis, a Subprogram of the Statistical Package for the Social SCiences (Nie, et al., 1975). In general, crosstabulation analysis computes and displays the frequency distribution of the number of cases that occurred for two or more discrete Variables. Data are displayed in a contigency table. 60 61 A 4 x 8 design (four tasks by eight groups) was programmed for crosstabulation analysis. The four tasks consisted of the following codes: I. ATASK l to ATASK 25 -- representing the task of superordination. 2. BTASK l to BTASK 25 -- representing the task of location. 3. CTASK l to CTASK 25 —- representing the task of similarity. 4. DTASK l to DTASK 25 -— representing the taSk of part. The numbers one to 25 corresponded to the 25 lexical stimulus items that were subjected to each task condition. The eight groups represented all possible combina- tions of the three fixed independent variables (age, sex and education) investigated in the present study. Each variable consisted of two levels. Table 6 lists the eight groups in— cluded in the crosstabulation analysis. The contingency tables obtained from the crosstabulation analysis provided the within cell (individual group) frequencies and the rOW'marginal frequencies. Of major importance to the present lhTVestigation were the row marginal frequencies because t1"lese values represented the combined frequency of occurrence SCores across all groups. The primary and common response measures were taken directly from the marginal row counts. 62 TABLE 6 LIST OF THE EIGHT GROUPS INCLUDED IN THE CROSSTABULATION ANALYSIS AS A FUNCTION OF THE SUBJECT'S AGE, SEX AND EDUCATION Group Agea Sexb EducationC Nd l -O F (HS 18 2 - M (HS 15 3 — F (HS 20 4 - M < HS 16 5 — F 3 HS 18 6 - M >HS 15 7 — F "’7Hs 15 8 M 3Hs 15 Notes: = Age levels consisted of Young—Old (Y—O) and Old—Old (O-O). b = Sex levels consisted of Females (F) and Males (M). C = Educational levels consisted of subjects with less than high school completion ( ( HS) and subjects with equal to or greater than high school completion ( 3 HS). d = The letter "N" represents the number of subjects within each group. The primary responses of each stimulus item and the Corresponding frequencies of occurrence are listed in Tables 7, 8, 9 and 10. The table numbers correspond to the four experimental tasks of superordination, similarity, location and part, respectively. The 25 lexical stimulus items are listed alphabetically in each table to provide for easier COmparison across tasks. As was alluded to in Chapter Three (Research Methodology), the actual ordinal presentation of ‘ . «u- B. 63 TABLE 7 PRIMARY RESPONSESAND CORRESPONDING FREQUENCIES OF OCCURRENCE (f) FOR SUPERORDINATION TASK Lexical Stimulus Item anger bread cabbage candle city eagle girl house knife memory moth music nurse rug soldier stomach street sun table tiger thief tobacco water whiskey wool Primary Response mad food vegetable light town bird female; feminine; femininity home tool mind insect sound helper cover; covering; floor covering fighter organ road; roadway light furniture animal robber smoke; smoking drink; drinking drink material [H1 39 75 126 102 52 129 74 35 36 25 89 30 20 26 31 36 28 113 118 37 59 46 48 28 PRIMARY 64 TABLE 8 RESPONSESAND CORRESPONDING FREQUENCIES OF OCCURRENCE (f) FOR SIMILARITY TASK Lexical Stimulus Item Primary Response f anger mad 40 bread cake 33 cabbage lettuce 46 candle light 43 city town 96 eagle bird 55 girl woman 47 house barn 28 knife sword 27 memory remember; remembering; remembrance 27 moth butterfly 52 music sing; singing 29 nurse doctor 57 rug carpet; carpeting 83 soldier sailor 42 stomach bellya 22 street road 57 sun moon 80 table stand 37 tiger lion 78 thief robber 61 tobacco cigarette 41 water milk 21 whiskey gin 25 wool cotton 41 Note: a = Belly and intestines were equal in frequency of occurrence (f = 22). The decision for selecting belly as the primary response was based on flipping a coin. 65 TABLE 9 PRIMARY RESPONSESAND CORRESPONDING FREQUENCIES OF OCCURRENCE (f) FOR LOCATION TASK Lexical Stimulus Item Primary Response f anger mind 42 bread bakery .28 cabbage garden 100 candle holder _ 28 city state 67 eagle sky 68 girl school 46 house street 38 knife kitchen 51 memory mind 62 moth clothes; clothing 50 music church 16 nurse hospital 131 rug floor 105 soldier army 99 stomach body 91 street city 80 sun sky 127 table kitchen 61 tiger jungle 48 thief jail 33 tobacco pipea 28 water lake 35 whiskey bottle 38 wool sheep 94 Note: a = Pipe and store were equal in frequency of occurrence (f = 28). The decision for selecting pipe as the primary response was based on flipping a coin. 66 TABLE 10 PRIMARY RESPONSESAND CORRESPONDING FREQUENCIES OF OCCURRENCE (f) FOR PART TASK Lexical Stimulus Item Primary Response f anger mad 30 bread flour 53 cabbage leaf lOl candle wick 53 city people 25 eagle wings; wing 74 girl hair 36 house room 30 knife blade 61 memory mind 24 moth wings 91 music notes 42 nurse uniform 57 rug wool 40 soldier uniform 48 stomach lining 30 street pavement 24 sun rays 41 table legs 106 tiger stripes 41 thief handsa 24 tobacco leaf 96 water liquid 32 whiskey alcohol 49 wool fiber 23 Note: a = Hands and gun were equal in frequency of occurrence (f = 24). The decision for selecting hands as the primary response was based on flipping a coin. 67 the 25 stimulus items varied from task to task. Pages three and four of the answer form (Appendix D) show the actual order used to present the stimulus items to the elder— ly subjects on a specific task. Incorporating the rules established by Laffal (1955), response words differing in form but coming under the follow- ing rules were combined as one type of response: 1. Singular and plural responses of the same word. 2. Responses containing the same root but differing only in terms of a suffix, prefix or infix that did not essentially alter the basic word. The frequency count of the primary responses, as elicited by the 132 elderly subjects, ranged from 16 to 131 on a specific stimulus item. The mean frequencies of the primary responses under each task condition appear in Table 11. The mean scores indicate that, on the average, the greatest amount of primary response agreement among the subjects was given when subjects were asked to name the location of the stimuli. The task having the next highest primary response agreement was superordination followed by part and similarity tasks, respectively. The second group of measures obtained from the row marginal counts were the frequencies of occurrence of the common responses elicited by the 132 elderly subjects. As was defined in the terminology section (Chapter One), a common response is obtained when two or more subjects give the same response to a lexical stimulus item. An arbitrary 68 TABLE 11 MEAN FREQUENCIES OF OCCURRENCE OF THE PRIMARY RESPONSES UNDER FOUR TASK CONDITIONS Task Condition Mean Frequency of Occurrence Superordination 58.12 Similarity 46.72 Location 62.64 Part 49.24 cut—off level of 85 percent of the total population (n = 112) was established for purposes of reporting the common responses obtained under the four experimental tasks. The common responses were selected on the basis of their corresponding frequencies of occurrence. Selection consiSted of ordinally choosing the highest to the lowest frequencies of occurrence until at least 85 percent of the population was represented. The common responses and the associated frequency distribu— tions for each task are reported in Appendices E, F, G and H. For purposes of comparison across tasks, the stimulus items and the common responses are listed alphabetically. The primary responses appearing in Tables 7 through 10 are in- cluded in the four Appendices. The primary responses are listed with the common responses for easier accessibility. A third descriptive measure obtained from the cross- tabulation analysis was the diversity score for each lexical stimulus item on each of the four experimental tasks. The 69 diversity score represents response variability or hetero— geneity. Diversity scores were calculated by counting the num- ber of different responses given to a lexical stimulus item under a specific task condition. The diversity scores for each of the stimuli appear in Table 12. Each "NR2" (indi— cating no response on a particular item after two attempts) was counted as an idiosyncratic response and was included in the diversity scores. Row means (indicating the average diversity score for each lexical stimulus item across tasks) and column means (indicating the average diversity for each task condition) are also provided. Results of the column means showed that the elderly subjects had the least amount of response diversity on the restricted association task of location and the greatest amount of diversity when responses were restricted to naming a part of the stimuli. Stated differently, the size of the associative networks for the stimulus items used in the present investigation appeared to be smallest for the task condition of location and largest for the task condition of part. The remaining two conditions (similarity and superordination) obtained diversity scores that fell between the above task conditions. To analyze the effects of age, sex and education on the diversity scores obtained, three two-level split group Comparisons were constructed. The analyses consisted of Comparing the mean diversity scores between the young—old (Y—O) Adam: ecmuoch.Hm om.e~ oH.eH Ne.mm em.mH cams Hmms oo.m~ me Hm Hm Hm H003 em.mH em mH em mH smmena oo.mm Hm mm mm mH “mph; m>.mH mH wH mN ma ooomnou om.mm em om em em Heru me.mH mm NH mH m HmeHu om.mH m HH mm OH mHnmu oo.mH mm 4 mH Hm cam me.mm em mH NH em ummuum om.mm em mH mm mH somsoum om.em mm m om mm hmHeHOm om.eH mm 4 mH NH may mm.mm mm N em oe whose mm.mm em em em em onss me.mH mH Hm mm 0H shoe mh.NN mN m vN mN wuoEwE me.mH m m Hm mH mHHcH me.mm mm mH em mH mmsor mm.em mm mm em NH HHHm mN.ma NH ma MN N macaw oo.mm mm mm eH mm muHo me.mH HH mm eH e 0Heqmo oo.NH HH mH GH e wmmnnmo mm.mH HH mm mm mH pawns om.mm mm em om om Human com: puma soHpmoOH hpmHmHHEHm QlowcHUHOmedm EwuH msasfimhm EmuH HMOwaH coHpHecoo ghee mZOHBHQZOU Mmdfi MDOM mmE MMQZD SHBH mDHDSHBm AHQ NH mqmdfi 71 and the Old-Old (O-O), between the females (F) and males (M) and between those subjects who did not complete high school (<.HS) and those subjects with equal to or greater than high school completion (3 HS). Table 13 lists the mean diversity scores for each group on each of the four experimental tasks. The row means and column means represent the group mean combined across tasks and the mean diversity score for each task across all subjects, respectively. The numerical value beside each group represents the number of subjects within the group. As was exemplified by the overall combined group mean diversity scores (Table 13), on the average the diversity score for the task of location was lowest, whereas the highest number of different responses was obtained under the part task condition. Figures 3, 4, 5, and 6 graphically represent the mean diversity scores for each of the two-level groups of age, sex and education under the four task conditions. The following trends were observed in the graphic analyses: 1. The old-old group had less diversity among the responses given for the superordination and similarity tasks. However, the trend was reversed for the location and part tasks. The young-old scored lower on the diversity measures for the latter two tasks. 2. The diversity scores for females were essentially the same as the male subjects on three of the four tasks (similarity, location and part). Males scored slightly higher than females on the superordination task. “Ame n av mHmEmm Ho 0p Hmswm m “mmSOHw xmm mm cvo .AHw H CV GAME H 2 “Ame u av coeumHmEoo Hoozom smegmtcmnu Hmummum comhdeEoo Hoosom 5mm: swap mmmH w mm “mdsouo sonwosUm n Q “Ame u av UHOIUHO u Ono “Ame u cv UHOImcsow u 01% “museum mod n m "mmpoz Home: ccmuwvmn.eH me.nH m>.HH mm.mH om.mH dmmz cEUHoo . . . . . He He vH Nm pH Ne HH Ne mH we mH 2 He oxmm em.¢H eN.mH Nm.HH Nm.mH mv.NH m o o o I o mw mH «H we mH Nm OH Nm 0H vo MH mmm. .2 mm QGOHDMODUH 7 oe.mH Nm.mH ev.NH oe.mH vv.eH Wmv me.eH ee.mH em.mH em.eH m~.mH eeouo mm mmm< mm.vH Nm.nH om.HH me.mH mN.eH one new: uumm coeumooq MMHHMHHEHm comumcflcuonmmsm msouw 30m HHHdm aoHuHecoo ghee Nmm QZd ZOHBflUDQm .mw¢,m0 ZOHBUZDm ¢ mfi mmmoom NBHmmm>HQ ZfimE m0 MMdEEDm ma mqmdfi Mean Diversity Score Task as Notes: 20 19 18 I7 16 73 _ Superordination - Diversity (HS Age Sex Education Figure 3. Mean Diversity Scores for Superordination a Function of Age, Sex and Education. a = Age Groups: Y-O = Young-Old; 0—0 = Old-Old U C) = Sex Groups: F = Female; M = Male = Education Groups; < HS = less than high school completion; ‘5 HS = Equal to or greater than high school completion 74 2o e in, "a rity - D iv e rs it Y I----’ 0 "— Y ‘1': 1| lii I. ‘l::’ ) pw' ho.- ::' _-- :::: n. .I ..... .. nnnnnnnn .. _ fl cccccccccc I. ......... m t ......... ».. ....... Io... ..... p ..... nu.- oooooooo .- |I .I-II’ ---------- .......... .......... ll:' .......... .......... .......... . .......... .......... I. .......... . ........ I: :1 ......... --. ......... ......... 1 ......... .......... .......... .......... l::: .......... .......... m 2 .......... .......... .......... a) L— .......... .......... .......... : oooooooooo .......... ‘l .......... IF: .......... r ‘1' .o. 7r . a .......... I _ :::::::':: 1 ''''''' ---------- ........... ‘l‘ nnnnnnnnnn P.0000oloui 1 4 1 _ ::::::::: ............ --------- ........... 4 p . -.. I'D-I nnnnnn .20. ,ogcogo-OIII aaaaaaa on pang-nuclei :CICOI: 4 — :::::::::: IOODIIOOIOI: -‘ :::::::::: ”uncau-I: :: .... - ' [I'IIIIIZI I """""" £22222 ::: ..I ‘ """ . . ... >.......oa a... r o::.:.: .. E22......Z {22322223221 1.. ........ _ t::::::::: ;':::::::::: ;:::.:::::. . b oooooo o 4 1 ooooooo an! ..... ;:::::.:': t ........... ‘:..:.: :..1 .......... {31:33:31 .._.".. .: u I 323223: 4 1 13323113. .......... b‘l""“"‘ ”use..." III 4 -:I::..'. J ........... ;:::::::::: 1 --------- : ....... .......'.' D'Il.’-.'.. ’.:l OOOOOOO ‘ p.n.unn:-o P :::::::.:: :::-::::::: a": n . L.u'-oo:ol ...QIIll.-. ::::::..:: 4 0.0 0 ' I P III-III... ........ — .uuu-"O ”nun-fl 4 ‘ .-.::-::::: p....o.--¢: ...uuoo-no- I .......... ........... p....cuoo-1 ............ p....oo--O1 ...n:':::: :‘IcouOOU': o-o-UO . 0-0“ . IN 1 ou::: :....o--Iu: uuuuuuuuu ........... ooooooooooo ........... ........... ----------- .......... t .......... nnnnnnnnnnn ........... ooooooooooo ........... DIG-...OOI: ......... uuuuuuuuuu ..... ..III‘ . oooooo 1" ooooooooo - INT"; ooooooo ooooooooooo .......... ........... ........... .......... .......... , li:;;;:;:: | III. D II.- I. L_‘ , A_ 9e, S, e x Educ a t. 'o n a s a F uF n i 0 g t u i r 0 e n 4 o f. A M geIan S D e i X V e a r n dsi Et dY US Gator i e O s n of o r S i m i l a r i t Y T a s k N O t e S a b = C=Age s e G X r G on r p 0 s u p: S = Y F o Fe:oun a 1 9- e 0 7 l M 5. ='O M '0 a 1 = e o 1 d '0 1d — E d C u some c a lrlfr;) :1. 1t; I 0 e10 0 t 1 in O G C°§£:9Pp > s e t H; i s o < n :: e23 a: 1 l tgss 0 t r h gin e h a i t g e h r ts hch a o n o h1 i 9h 75 20 __ Location- Diversity 18— an 11— Mean Diversny Score :3 I Age Sex Education Figure 5. Mean Diversity Scores for Location Task as a Function of Age, Sex and Education. Notes: a = Age Groups: Y-O = Young-Old; 0-0 = Old—Old b = Sex Groups: F = Female; M = Male = Education Groups; ( HS = less than high school completion; ‘3 HS = equal to or greater than high school completion ‘ Mean Diversity Score Notes: 76 Part - Diversity '9 ()-C) ..... -— I‘l‘I'I'I‘ Y- o .......... Z'I'I-I'I' — ':':':':‘: }.0.I.I:. lllllllllll ’:.:I .............. *.'.'.'.. I-I'Z'I'Z' .................... I] -— .................... ..................... - ..................... z':-:-:-:- '-:-:-:-:-: .......... {3.333. 'I'Z'I'I': ’I'I'I'I'I' ' :-:-:-:-:- . . . . 1 {3.3.3. - : : : :.:« :-:-:-:-:-: C ttttt ......... I >.'.°.'.'.' ..................... ''''''''''''''''''' ....... j p . . . . . ...:._.-. .:.:.:.:.:. 33.31 :.;.;.;.;.: .......... ...................... ........... .......... 15 — ........ J ........... ........... - ........ 1 »:-:-:-:-:- .................... ..................... ........... '1'3'3'1'. 'I'I-I'I'Z' ..................... :-:-:-:-:« :-:-:-:-:-: .................... 14 .. .................... ...... 1 O O I I I ..:.:.:..4 .:.:.:.:.: ......... ................... .......... ......... ................... ................... .......... ................... ................. ........... '3 — ''''''''''''''''''' ................... ''''''''''''''''''' ''''''''''''''''''' :-:-:-:-: -:-:-:-:-: ................... ......... ................... ................... ................. .............. — ................... ...... ..'.‘.'. I ..... .. I I i .......... ................... .......... ......... 1'.'.'.'.'. ............. II... I O .‘D‘. I’ll ............ IIIIIIIIIIIII * ................... I _ ..‘..._.. . ......... 3.1.5:.- .:.:.:.;. '..'.-.": :.:.:.:.: :.:.:.:.:. .‘O'l'l’. ':l ....... .................. .................. ............ :-:.:-:-: _ Q....'... .:I ....... 0:..I'l.l D.I.c:.:.: ................... ................ ............. 'J ..I.l-l.. . ......... E ................. 4 ............. 'J .............. ............... .................... — LI 3 o n o :I ........ ................... A ........ :. :..I:l:.:. }.o o. ‘‘‘‘‘‘‘‘‘‘ .................. ». . fl ........ D i I I. ........ u o o n .......... :.:.:.:. .......... ........ .......... . . . ...... 8 C I I ..i . . . . ' ‘ ' :.‘ .o..:.:.: I I I .'j . C I ... ..... A A ‘-.'......l ‘ ............... :-:-:‘:-:' .- ------- 0000000 ’3‘ ' '_'_0.I 0 -:-:-:.: '''''' Age Sex Education Figure 6. Mean Diversity Scores for Part Task as a Function of b Age, Sex and Education. Age Groups: Y-O = Young-Old; 0—0 = Old-Old Sex Groups: .F = Female; M= Male Education Gropps; < HS = less than high school completion; ) HS = equal to or greater than high . school completion 77 3. On the average, the elderly subjects with equal to or greater than high school education had lower diversity scores on all four of the experimental tasks, although the gap between the two groups narrowed considerably on the similarity task. Correlational Measures of Diversity Scores as a Function of Age, Sex and Education To determine whether the pattern of scores for a set of stimulus words under a given task condition was influenced by the subjects' age, sex, or education, correlational analyses were calculated. Employing an item-by-item analysis of the diversity scores obtained on each stimulus item across tasks, the Pearson product moment correlational procedure was used to find the relationship between the diversity scores of the Y-O and 0-0 age groups, the F and M sex groups, and the (,HS and 1; HS education groups for all tasks. The results of the correlational calculations appear in Table 14. The compara- tively high correlational scores depicted in Table 14 indicate that the lexical stimuli formed a relatively consistent scale of agreement or disagreement across the tasks regardless of the subject group comparison. Subject variation changed the mean diversity scores obtained on each of the items but had no effect on the relative ease or difficulty of each item. In summary, in the sample of stimulus items used, the tendency to agree was far more closely related to the nature of the stimulus item than to differences in subject groups. 78 TABLE 14 SUMMARY OF CORRELATION VALUES OF THE DIVERSITY SCORES OF 25 LEXICAL STIMULI AS A FUNCTION OF AGE, SEX AND EDUCATION Task Condition Group Superordination Similarity’ Location Part Y-O: 0—0a r = .98 r = .97 r = .95 r = .97 F: Mb .97 .98 .98 .98 (113:3 115C .97 .98 .97 .96 Notes: a Y-O = Young-Old; 0-0 = Old-Old b F = Females; M = Males C (HS Less than high school completion; “$HS = Equal to or greater than high school completion Frequency Distribution of Primary Response Scores The number of primary responses elicited by each of the 132 elderly subjects was determined by counting those responses that were identical to the primary responses obtained from the entire population samples (Tables 7, 8, 9 and 10). Since 25 primary responses were possible under each task condition a subject's primary response score could range from zero to 25. Thus, each subject had four scores (range = zero to 25) which represented the number of primary responses elicited by the subject under each of the four task conditions. To obtain a frequency distribution of primary response scores, the scores were subjected to percentile rank analysis. The rank order analysis was based on charting the distribution 79 of the lowest to the highest primary response scores obtained on each task. Tables 15, l6, l7 and 18 provide information on the frequency of occurrence of a given primary response score. The cumulative frequency distribution and cumulative percentage distribution of the scores are also included. TABLE 15 CUMULATIVE FREQUENCY SUPERORDINATE Cumulative Cumulative Raw Score Frequency Frequency Percentage 6 1 l (l % 7 7 8 6 8 12 20 15 9 18 38 29 10 21 59 45 ll 21 80 61 12 19 99 75 13 10 109 83 14 11 120 91 i 15 7 127 96 \ l6 3 130 99 l 17 2 132 100 80 TABLE 16 CUMULATIVE FREQUENCY SIMILARITY Cumulative Cumulative Raw Score Frequency Frequency Percentage 4 6 6 5 % 5 8 14 ll 6 12 26 20 7 15 41 31 8 18 59 45 9 19 78 59 10 21 99 75 ll 14 113 86 12 8 121 92 13 6 127 96 14 l 128 97 15 4 132 100 TABLE 17 CUMULATIVE FREQUENCY LOCATION Cumulative Cumulative Raw Score Frequency Frequency Percentage 7 2 2 2 % 8 7 9 7 9 15 24 18 10 15 39 30 ll 20 59 45 12 27 86 65 l3 14 100 76 l4 14 114 87 15 7 121 92 16 6 127 96 17 2 129 98 18 3 132 100 81 TABLE 18 CUMULATIVE FREQUENCY PART Cumulative Cumulative Raw Score Frequency Frequency Percentage 4 3 3 2 % 5 6 9 7 6 9 18 14 7 13 31 24 8 14 45 34 9 24 69 52 10 21 90 68 11 15 105 80 12 14 119 90 13 6 125 95 14 3 128 97 15 3 131 99 18 1 132 100 Analyses of Variance of Primary Response Scores as a Function of Age, Sex and Education The purpose of the analyses of variance was to determine whether for any given task, a difference existed in the number of primary responses elicited by the subjects as a function of age, sex or education. The mean primary response scores for each split group analysis appear in Table 19. Row means and column means are included. The statistical analyses employed in the investiga— tion were three two-way, fixed effects analyses of variances (2 x 4) with repeated observations. The first analysis of variance was calculated to analyze the primary response scores Cflotained by the subjects as a function of age (young—old, old—old) 82 va.oa hm.oa om.m hm.oa HN.OH mN.0H com: om.m mm.HH mw.m mm.OH com: cEsHoo Nm.m mm.NH Hm.w me.0H mm.P mm.m mv.HH mH.m 0N.HH mmv coflumoscm wN.m mm.HH Nm.m mm.0H and: CESHOO em.m om.HH vw.m mh.0H mHoz Nm.m Nm.HH om.m mN.HH mHmEom xom Nm.m mm.HH em.m oo.HH and: GESHOU Nm.m me.HH he.m mm.OH CHOIUHO me.m ee.HH mm.m em.HH eHoumcsos mod pumm QOHHMOOH upHHmHHEHm compmchuonmmsm muonw wHHmm QHMOHWOHMHmcH mamoflmoq somumccoo xmne ZOHBHO mmmzommmm wm Zfimz m0 WMfiSZDm ma mqmflfi 83 Table 19 (cont'd.) Notes: aGrand Means: Logical Tasks Age = 9.92 Sex = 9.90 Education = 9.91 bGrand Means: Infralogical Tasks Age = 10.64 Sex = 10.57 Education = 10.59 84 TABLE 20. TWO WAY FIXED EFFECTS (AGE x TASK) ANALYSIS OF VARIANCE 'WITH REPEATED MEASURES AND EQUAL GROUP SIZE FOR THE NUMBER OF PRIMARY) RESPONSES ELICITED BY SUBJECTS Source of Variation Degrees of Mean Fa Probability ’ Freedom Square Value of Statistic A (Age) 1 0.61 0.07 0.7856 Error (Age) 130 8.26 B (Task) 3 219.78 39.96 0.0001* AB (Age x Task) 3 12.34 2.24 0.0864 Error (Within) 390 5.50 * Significant at .01 level of confidence. Notes: a = Critical F values at .01 level: Factor Critical F Value Age 6.63 Task 3.78 Age by Task 3.78 85 and task condition (superordination, similarity, location and part). The second analysis of variance was calculated to analyze the primary response scores as a function of sex (male, female) and task condition. The third analysis of variance was calculated to test statistically the primary response scores as a function of education (less than high school completion, equal to or greater than high school completion) and task condition. Primary Responses as a Function of Age and Task Condition Table 20 represents a two-way, fixed effects (age by task) analysis of variance with repeated observations and equal group size (Young—Old = 66; Old—Old = 66). The results indicate that there is no significant main effect for age (factor A) and no significant age by task interaction (factor AB). However, the task main effect (factor B) was statistically significant at the .05 level. Three post hoc planned comparisons for the task main effect were developed to test for the following paired contrasts: Contrast One. To determine whether a significant difference existed between the two logical tasks (super- ordination and similarity) on the primary response scores obtained by the subjects. Contrast Two. To determine whether a significant difference existed between the two infralogical tasks (lOcation and part)on the primary response scores obtained. 86 Contrast Three. To determine whether a significant difference existed between the combined primary response scores for the two logical tasks and the combined primary response scores for the two infralogical tasks. The planned comparison post hoc procedure was used to eliminate the pro— blem of redundancy that is inherent to repeated measures designs. Table 21 represents the results of the three post hoc planned comparisons. When task conditions were statis- tically analyzed for the pooled age groups, all three contrasts were found to be significant at the .01 level. Specifically: a) When age groups are combined, a statistically significant difference exists between the primary response scores obtained on the superordination task and the similarity task. The group mean for superordination was 11.00 versus 8.84 on the similarity task. b) A statistically significant difference exists between the primary response scores obtained on the location task and the part task for the combined age groups. The group mean for location was 11.86 versus 9.32 on the part task. c) A statistically significant difference exists between primary response scores obtained on the logical tasks and the infralogical tasks. The group grand mean for the logical tasks was 9.92 versus a grand mean of 10.64 for the infralogical tasks. All mean values appear in Table 19. 87 TABLE 21 PLANNED COMPARISON POST HOC ANALYSIS OF THE TASK MAIN EFFECT FOR ALL SUBJECTS d Source of Variation Degrees of Mean F Probability Freedom Square Value of Stat1st1c Sup — Sima 1 307.58 59.29 0.0001* Error (Sup — Sim) 130 5.19 Loc — Prtb 1 399.97 65.08 0.0001* Error (Loc - Prt) 130 6.15 Log — Infc 1 65.52 12.68 0.0006* Error (Log — Inf) 130 5.17 * Significant at .01 level. Notes: a (Sup — Sim) Similarity. b (Loc — Prt) (Log - Inf) Contrast between Contrast between Contrast between (superordination and similarity) tasks (location and part). Critical F value at trasts = 6.63. superordination and location and part. logical tasks and infralogical .01 level for all three con— 88 Primary Responses as a Function of Sex and Task Condition Table 22 represents a two—way fixed effects analysis of variance (sex by task condition) with repeated measures and unequal group size (males = 61; females = 71). Factor A, the sex main effect, was found to be significant. Results indicate that, on the average, there is a significant difference between males and females on the number of primary responses elicited across task conditions. The task main effect (factor B) was also found to be significant. Sex by task interaction (factor AB), however, was not significant. The mean data for the primary response scores tabu— lated by sex and task condition appear in Table 19. The row mean for women (across all task conditions) was 10.57 primary responses; the row mean for men was 9.90. Results indicate that the female subjects had more primary responses than did their male counterparts. The analysis of variance found this difference to be significant. Figure 7 graphically traces the mean scores for primary responses as a function of sex across all task conditions. The planned comparisons contrasts for post hoc analysis Of the significant task main effect (factor B) appear in Table 23. When task conditions were statistically analyzed for the combined sex groups, all three contrasts were also found to be significant at the .01 level. Specifically: a) When sex groups were pooled, a statistically Significant difference was found between primary response scores obtained on superordination task and the similarity task. The group mean for superordination was 10.99 versus 8.82 on the Similarity task. i 89 TABLE 22 TWO-WAY FIXED EFFECT (SEX BY TASK) ANALYSIS OF VARIANCE WITH REPEATED MEASURES AND UNEQUAL GROUP SIZE FOR THE NUMBER OF PRIMARY RESPONSES ELICITED BY SUBJECTS Source of Variation Degrees of Mean Fa Probability Freedom Square Value of Statistic 1 \ A(Sex) l 62.90 8.12 0.0051* Error (Sex) 130 7.74 B(Task) 3 223.23 40.44 0.0001* ‘ AB(Sex by Task) 3 8.90 1.61 0.1899 Error (Within) 390 5.52 * Significant at the .01 level. Note: a Critical F values at .01 level: Factor Critical F Value Sex 6.63 Task 3.78 Sex by Task 3.78 A l 90 SEX 19 ~k female I n1a|e Mean Number of Primary Responses I 0 ’r l I l 1 super- similarity location part ordination Figure 7. Mean Primary Response Scores for Male and Female Subjects Under the Four Task Conditions. 91 TABLE 23 PLANNED COMPARISON POST HOC ANALYSIS OF THE TASK MAIN EFFECT FOR ALL SUBJECTS . . Degrees of Mean Fd Probability Source Of Variation Freedom Square Value of Statistic Sup — Sima 1 309.74 59.23 0.0001* Error (Sup - Sim) 130 5.23 Loc — Prtb 1 399.97 66.30 0.0001* Error (Loc — Prt) 130 6.03 Log - Info 1 64.82 12.21 0.0007* Error (Log — Inf) 130 5.31 * Significant at .01 level. Notes: a (Sup — Sim) = Contrast between superordination and similarity. b (Loc - Prt) = Contrast between location and part. (Log — Inf) = Contrast between logical tasks (superordination, similarity) and infralogical tasks (location and part). d Critical F value at .01 level for all three con— trasts = 6.63. ain- 92 b) A statistically significant difference was found between the primary response scores obtained on the location task and the part task for the combined sexes. Group means for the location and part tasks were 11.86 and 9.28, respectively. C) A statistically significant difference exists between primary response scores obtained on the logical tasks and the infralogical tasks. The group grand mean for the logical task was 9.90 versus a grand mean of 10.57 for the infralogical tasks. Primary Responses as a Function of Education and Task Condition The summary of the third two—way fixed effects (education by task) analysis of variance with repeated measures and unequal group size (less than high school completion = 63; equal to or greater than high school completion = 69) appears in Table 24. The education by task interaction (factor AB) was significant at the .05 level. The task main effect (factor B) was significant at the .01 level. The group main effect (factor A) was not significant. Results shown in Table 24 indicate that, on the average, education does not have a significant effect on the number of primary responses elicited by the subjects across the four task conditions. However, when a specific task condition is subjected to the two levels of education, a significant difference exists in the performance of the two groups in terms of the number of primary responses elicited. 93 TABLE 24 TWO-WAY FIXED EFFECTS (EDUCATION BY TASK) ANALYSIS OF VARIANCE WITH REPEATED MEASURES AND UNEQUAL GROUP SIZE FOR THE NUMBER OF PRIMARY RESPONSES ELICITED BY SUBJECTS a Source 0f Variation 3:23:51 °f £3212. vine ZE°§i§iiifiiC A (Education) 1 4.99 0.61 0.4375 Error (Education) 130 8.22 B(Task) 3 225.58 41.36 0.0001* AB(Education by Task) 3 16.61 3.05 0.0313** Error (Within) 390 5.45 i * Significant at .01 level ** Significant at .05 level. Note: a Critical F value at .01 and .05 level. Critical F Value Critical F Value Factor .01 level Education 6.63 Task 3.78 Education by Task 3.78 .05 level 3.84 2.60 2.60 94 Using the mean primary response scores for education (Table 19), a graphic representation of the education by task interaction was plotted (Figure 8). Noting the distance between mean points on a specific task, post-hoc t—test analyses with pooled variances were calculated. The first t-test contrast consisted of statistically testing the difference between the mean score of the (HS group and the mean score of the75HS group on the task of location. The results of the post—hoc analysis indicated that there was no significant difference between the two education groups at the .05 level. Since the numerical difference between the means of the remaining task conditions (superordination, similarity and part) was smaller than the contrast between the two means obtained on the location task and since the pooled variance was the same across tasks, it was assumed that the t—test contrasts of the other task conditions would not be statistically significant. Therefore, additional t—test analyses were not performed. One possible explanation for the significant inter- action effect was that the subjects with less than high school education had more primary response agreement than subjects with equal to or greater than high school education on three of the four tasks (superordination, similarity and part). On the task of location, those subjects with equal to or greater than high school education scored almost one point higher (.090)than the subjects not completing high school. 95 EDUCATION * less than high school education - equal or greater than high school 18 - education Mean Number of Primary Responses 0 ’f’ l l l 1 super- similarity location part ordination Figure 8. Interaction Between Education and Task Condition. 96 Discussion In this section, the results as decribed above are compared and contrasted to the findings of previous investigations. For purposes of clarity, the research questions are restated. Following the discussion of each question, clinical implications are delineated. Common Responses Elicited by the Elderly Subjects Research Question One. In the restricted associa— tion task of superordination, what are the frequencies of occurrence of the common responses given by 85 percent of the nonpathological elderly group sampled? A common response was defined as a response given by two or more subjects to a stimulus item. The common responses and corresponding frequencies of occurrence elicited by 85 percent of the elderly subjects (n = 112) on the superordination task appear in Appendix E (pp. 135-146 ). The frequency distribution of the common responses ranged from 129 to two with a mean frequency of 17.17. The mean diversity score was 19.36. Research Question Two. In the restricted association task of similarity, what are the frequencies of occurrence of the common responses given by 85 percent of the nonpathological elderly group sampled? The common responses and corresponding frequencies of occurrence elicited by at least 112 elderly subjects on the similarity task appear in Appendix F (pp. 147—161 )- The frequency distribution of the common responses ranged from 96 to two with a mean frequency of 12.13. The mean diversity score was 23.72. 97 Research Question Three. In the restricted associa— tion task of location, what are the frequencies of occurrence of the common responses given by 85 percent of the non- pathological elderly group sampled? Appendix G (pp.152—171 ) denoted the common re— sponses and corresponding frequencies of occurrence for the location task. The frequency distribution of the common responses ranged from 131 to two with a mean frequency of 18.50. The mean diversity score was 16.16. Research Question Four. In the restricted associa— tion task of part, what are the frequencies of occurrence of the common responses given by 85 percent of the non— pathological elderly group sampled? Appendix H (pp. 172-189 ) denotes the common re- sponses and corresponding frequencies of occurrence for the part task. The frequency distribution ranged from 106 to two with a mean frequency of 10.29. The mean diversity score was 26.80. An overview of the common responses and associated frequency distributions indicates that the greatest amount of response agreement occurred on the location task. The part task had the least amount of response agreement. Clinically, the data provided in Appendices E, F, G, and H could be useful when comparing "normal" elderly persons with elderly individuals whose language abilities have been disrupted by a neuropathology. With normative data available, reductions in response agreement and changes in normally elicited responses could be more easily detected. 98 Primary Responses Elicited by the Elderly Subjects Research Question Five. In the restricted associa— tion tasks of superordination, similarity, location and part, what primary responses occur for each of the 25 lexical stimulus items? A primary response was defined as the most frequently occurring single response given by the subjects to each test stimulus item. The primary responses and corresponding frequencies of occurrence for the four tasks appear in Tables 7, 8, 9, and 10 (pp. 63 - 66). A comparison of the primary responses elicited by the 100 college students used as subjects to Obtain the Michigan Restricted Assgciation Norms (Riegel, 1965) with the primary responses elicited by the 132 elderly subjects of the present investigation Showed the following results: 1. Twelve of the 25 primary responses elicited by the college sample were identical to the primary responses elicited by the elderly subjects on the superordination task. The mean frequency of occurrence was 50.24 for the college subjects versus 58.12 for the elderly sample. 2. Fourteen of the 25 primary responses elicited by the college sample were identical to the primary responses elicited by the elderly subjects on the similarity task. The mean frequency of occurrence was 37.16 for the college group versus 46.72 for the elderly subjects. 3. Twenty of the 25 primary responses elicited by the college sample were identical to the primary responses elicited by the elderly group on the location task. The 99 mean frequency of occurrence was 42.04 for the college group versus 62.64 for the elderly subjects. 4. Fifteen of the 25 primary responses elicited by the college sample were identical to the primary responses elicited by the elderly group on the part task. The mean frequency of occurrence was 33.16 for the college students versus 49.24 for the elderly subjects. The primary responses that were identical to those elicited by subjects used to obtain the Michigan Restricted Association Norms (Riegel, 1965) appear in Table 25. Four of the stimulus items (cabbage, rug, tiger, and wool) had matching primary responses on all four tasks. Six stimulus items (candle, eagle, moth, table, tobacco, and whiskey) had matching primary responses on three of the tasks. The location task obtained the highest number of identical primary responses for both groups of subjects. Clinically, the identical primary responses elicited by the college sample and the elderly sample indicate that a certain amount of consistency is maintained across generations on the four restricted association tasks investigated in the study. The consistency of primary responses supports the theory that tests measuring crystallized intelligence (which includes stored verbal information) exhibit the least differ- ences between young and old subjects (Labouvie-Vief, 1976). TABLE 25 PRIMARY RESPONSES IDENTICAL TO THOSE APPEARING NORMS (RIEGEL, 1965) IN THE MICHIGAN RESTRICTED ASSOCIATION Stimulus Item anger bread cabbage candle city eagle girl house knife memory moth music nurse rug soldier stomach street sun table thief tiger tobacco water whiskey wool TOTAL Task Condition Superordination Similarity Location Part mad mind food flour vegetable lettuce garden leaf light light holder town state bird sky wings woman room kitchen blade mind insect butterfly wing sound note doctor hospital covering carpet floor wool army uniform organ intestinea body road city pavement sky furniture kitchen leg robber jail animal lion jungle stripe cigarette pipe leaf lake drink bottle alcohol material cotton sheep fiber 12 14 20 15 101 Table 25 (cont'd.) a Note: = Intestines and belly were equal in frequency distibution in the present investigation. Belly was selected as the primary response on the basis of flipping a coin. 102 The Effects of Age on Primary Response Groups Research Question Six. Is there a significant difference between subjects, ages 65 to 74 years and subjects, ages 75 years and above, on restricted associa— tion tasks of superordination, similarity, location and part as measured by the number of primary responses given by each subject in each age group? Results of the data indicated that there was no statistically significant difference between the young— old and old-old subjects on the primary response scores obtained across the four tasks. To date, only one investi— gation (Riegel and Riegel, 1964) has attempted to find differences among elderly subjects as a function of age. The investigation divided 76 elderly subjects into five age categories (55-59, 60-64, 70-74, and 75 years and over). Comparing the five elderly groups with 120 young subjects (ages 16—20 years). Riegel and Riegel found that as age continued to increase, primary response scores decreased but the decrease was not significant. The results of the present study are somewhat jxiagreement with the findings of the Riegels' investigatfixnh Statistically, both age groups appeared to perfornl the same across tasks. However, the results depicted in Fi4§ure 3 showed that primary response agreement betweerl the two age groups changed as a function of task condijfirwn The young-old group had less agreement than the old-old subjects on the logical tasks (superordination, similarity) but tended to have 103 more response agreement than the old—old on the infra- logical tasks. Although the results of two studies are comparable, a caution must be made with respect to the manner in which data were collected. The Riegel study consisted of five grammatical categories (concrete nouns, abstract nouns, verbs, adjectives and'adverbs). Only nouns were used as stimuli in this investigation. In addition, stimulus items in the Riegel study were presented orally instead of in a standard audio-visual procedure. Clinically, the results of this investigation imply that age will have no significant effect on the amount of response agreement or response variability that occurs between, for example, two groups of elderly aphasic patients. However, a fluctuating pattern of agreement may occur depending upon the Specific task conditions being investigated. Future research should compare other age groups (e.g., young adult, middle age) with nonpathological elderly groups of subjects. Perhaps a sequential research design will be desirable, whereby the various age groups are investigated at designated intervals across a given time span. With this type of design, changes in response agreement or disagreement (diversity) across groups and within individuals can be detected simultaneously. In 104 addition, perhaps a five to ten year gap between the age groups (e.g., 25-35, 45—55, 65—75, etc.) should be in- cluded in the design. Leaving spans of time between age groups could possibly make any age differences in response agreement or response diversity more pronounced. The Effects of Sex on Primary Response Scores Research Question Seven. IS there a significant difference between male and female subjects on restricted association tasks of superordination, similarity, location and part as measured by the number of primary responses given by each subject in each sex group? The sex main effect was found to be significant at the .01 level of confidence. The results indicate that women achieved higher primary response scores than men across all task conditions. Studies by Palermo (1963) and Palermo and Jenkins (1965) showed Similar findings, although in the latter two investigations male subjects obtained higher scores than females on the superordination task. When these results are applied to a rehabilitation Situation one can only speculate that primary response scores of aphasic men and women should tend to be different with any variance usually favoring the women. The reason for this difference is as yet unexplained. lHopefully, future studies will investigate differences between male and female aphasic 105 patients to see if the statistically significant sex main effect found in nonpathological elderly subjects is upheld in pathological (i.e., aphasic) groups of elderly subjects. The Effects of Education on Primary Response Scores Research Question Eight. Is there a significant difference between the subjects having less than high school education and subjects having equal to or greater than high school education on restricted association tasks of super— ordination, similarity, location and part as measured by the number of primary responses given by each subject in each education group? A significant interaction was found between educa— tion and the four task conditions. However, post—hoe t—test analyses, which contrasted the difference between the mean primary response scores of the two education groups on each task condition, revealed differences that were nonsignificant at the .05 level. An extrapolation of the t—test value ob— tained on the location task (t = 1.58 with 130 degrees of freedom) found the value to be statistically significant at the .062 level. The critical value of t at the .05 level was 1.64. Subjects with ) HS education had higher response agreement than subjects with ( HS education on the location task. However, more response agreement was found among Subjects with less than high school education on the remaining three tasks (superordination, similarity and part). Evidence of educational differences as reported in previous investigations is inconclusive. Herr (1950) 106 stated that an increase in education yielded an increase in response agreement. In contrast, Becher (1960) found that an increase in education led to a decrease in common responses. Sefer and Henrikson (1966) and Rosenzweig (1964) merely speculated that differences in response agreement existed as a function of years of formal education. No formal testing was conducted by the latter three investigators. When applied to a rehabilitative setting, the results of this study indicate that primary responses and common responses may vary as a function of the subject‘s years of formal education and the specific task condition being inves— tigated. Future research should investigate the effects of years of formal education on response agreement or response variability among a larger sample of nonpathological elderly individuals. Persons with grade school—, high school— and college-education should be included as subjects. In addi— tion, since aphasia was defined as a reduction in language (Schuell and Jenkins, 1961), a comparison between response agreement among elderly aphasics with less than high school completion and response agreement among elderly aphasics with equal to or greater than high school completion would pro- vide information on whether a neuropathological condition reduced the vocabulary of both education groups to the same level, irrespective of formal schooling. Closely related to years of formal schooling are life—long occupations held by elderly subjects. Occasionally, 107 occupational jargon influenced the responses elicited by some elderly subjects. For example, the life—long occupation of one male subject was an oil well driller. When asked to give a word that was similar to water, he responded with 011 because, according to him, "both are natural resources that are deep in the ground." When asked to give a synonym for stomach, a retired agriculturist whose area of concentration was animal husband- ry, responded with "crop". He stated that "It's the name we use for a Chicken's stomach." Clinically, occupational jargon may be useful stimuli for prompting aphasic patients to elicit a target word. Future investigations should employ occupational jargon as stimuli and have subjects respond with "a word that would be given by most people in your profession." The Effects of Task Condition on Primary Response Scores Research Question Nine. Does the pattern of restricted association responses differ significantly between the logical instructions (superordination, similarity) and infralogical instructions (location, part) as a function of age, sex or education? Two types of tasks were investigated, logical tasks (superordination, similarity) and infralogical tasks (location, part). Post-hoc comparisons were calculated to test significant task main effects across age, sex and edu- cation. Three contrasts were constructed to a) compare differences within the task-type (superordination minus similarity; location minus part) and b) compare differences 108 between the task—type (combined logical tasks minus combined infralogical tasks). When the three contrasts were analyzed across age and sex, all three post—hoc analyses were statis— tically significant at the .01 level. Although the task main effect was statistically significant at the .01 level across education, the significant interaction between education and task condition warranted no further discussion of the task main effect. Based on the column means appearing in Table 19, the rank order of the four task conditions for primary re- sponse agreement is as follows: 1. Infralogical task of location 2. Logical task of superordination 3. Infralogical task of part 4. Logical task of Similarity Stated differently, for the elderly population sampled, the highest amount of primary response agreement occurred on the infralogical task of location, whereas the greatest amount of response diversity occurred on the logical task of similarity. This pattern was consistent across age, sex and education. Comparing the results of this study to those obtained by Riegel and Riegel (1964), it would appear that using location responses as clue words would increase a subject's chances for carrectly identifying a target word. Riegel and Riegel fOUIMi similars to be the most efficient single clue bUt. in this study, similars had the least amount of response agrooment among the elderly subjects. In addition, the rank order analyses implies that mixed clues 109 (e.g., location and superordination) would produce more correct identification of the target words than double logical (e.g., superordination and similarity) and double infralogical (e.g., location and part) clues. Clinically, the results indicate that determining the location of an object may, perhaps, be the easiest task for the aphasic patient to perform. Rehabilitation specialists should begin with the location task and present more diffiCult levels of association in the order indicated. Future studies should investigate other logical and infralogical tasks that were used to collect the Michigan Restricted Association Norms (Riegel, 1965). Meanwhile, responses appearing in Appendices E, F, G and H can be used as clue words to investigate whether the highest amount of correct identification of the original stimulus items can be attributed to single—, double—, logical—, infralogical-, or mixed clues. CHAPTER V SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Summary The major objectives of this investigation were (1) to provide data on the responses elicited by non— pathological elderly individuals on four tasks of restricted word association and (2) to determine whether or not differences in the responses elicited could be attributed to the age, sex or education of the elderly persons. The research procedure consisted of (1) an assess— ment of the subject's hearing, vision and mental ability and (2) a test session. One hundred thirty-two elderly individuals, who passed the screening requirements, served as subjects. Each subject was seen individually for approxi— mately 40 minutes. A synchronized audio-visual presentation of 25 words was used to elicit responses to four restricted word association tasks. The task conditions consisted of two logical tasks (superordination, similarity) and two infra— logical tasks (location, part). The same set of 25 words 110 111 as presented on each task. The order of stimuli presen- tation varied from task to task depending on the rank order of the "easiest" to "least easiest" stimulus item on each task. Data from the test procedure were analyzed employing descriptive and inferential statistics. Descriptive statistical analyses were used to: 1. determine the common responses and corresponding frequencies of occurrence elicited by 85 percent of the total population sampled on the four tasks; 2. determine the primary responses and corresponding frequencies of occurrence elicited by the total population sampled on the four tasks; 3. determine the diversity scores for each lexical stimulus item on the four tasks. Results indicated that the location task had the greatest amount of response agreement, whereas the similarity task had the greatest amount of response diversity. Three two-way, fixed effects analyses of variance were employed. The first analysis of variance was used to investigate the difference in primary response scores as a function of age and task condition. A non-significant difference was found between the two age groups. The inter- action between eige and task was also non-significant. The differences betnneen the three planned compariSOn contrasts for post hOC analyses of the statistically significant task main effect Were all significant at the .05 level. 112 The second analysis of variance investigated the difference in primary response scores as a function of sex and task condition. Females were found to have significantly more response agreement than their male counterparts. Post— hoc analysis of the significant task main effect showed all three contrasts to be significant. The final analysis of variance explored differences in primary response scores as a function of education and task condition. A significant interaction between education and task condition was found. Post—hoc t—test analyses yielded non—significant results, although the difference between the two education groups on the task of location was significant at the .062 level. Conclusions The results and statistical interpretation of the word association data elicited by the 132 elderly subjects suggest the follOWing conclusions: 1- The use of a remotely controlled, synchronized audio—visual procedure is advantageous in minimizing mis— interpretation of the stimulus signals by the elderly subjects. 2. Data have been obtained on restricted word associations of elderly subjects on the tasks of super— ordination, SjJnilarity, location and part. The collected data Will be Useful for comparison with future investigations 0f language patterns of the elderly. In addition, the data 113 can be used as a rehabilitative tool with elderly aphasic patients exhibiting semantic confusions. 3. The infralogical location task is the "easiest" task for the elderly as determined by overall response agreement. The logical similarity task is the "least easiest" for the elderly as determined by the diversity scores. This pattern is consistent regardless of the age, sex or educational level of the subject. The fact that the location task proved to be the "easiest" task for the elderly subjects may be influenced by the specific nouns selected as stimuli which may have a commonly recognized place in the environment. In View of the present findings, however, habilitation of neuropathological geriatric patients ex— hibiting semantic confusions should begin with a location task followed by a superordination task. Part and similarity tasks should be introduced last. 4. The fact that overall primary response agree— ment was unaffected by the age of the elderly subjects indicates that restricted word association responses remain relatively unchanged for elderly persons aged 65 years and above. 5- Female elderly subjects have significantly more response agreement than male subjects. In speculation, the majority Of‘ the females were housewives whereas the males occupied a variety-of employment positions. Perhaps the "occupational Luiiformity" of the females and the "occupational diversity" of the males influenced the results obtained. In addition, the sex of the investigator (female) may have been 114 a contributing factor to the higher response agreement elicited by the elderly females. 6. The number of formal years of schooling in— fluences the response agreement obtained by elderly subjects. Results suggest that persons with more years of formal schooling are exposed to more words in the language and thus, have a larger semantic domain. The larger semantic domain allows the person to have a variety of word associa— tion options from which to select. Therefore, less response agreement occurs. The fact that elderly subjects with a greater number of years of formal schooling obtained a higher common response score on the location task than subjects with less years of formal schooling may have been a factor of the words used as stimuli. Recommendations The following are suggestions for continued research exploration: 1. Future studies should employ equal cell sizes for the age, sex and education variables. A concomitant suggestion would be to enlarge cell size well beyond the 15 subjects. Both of the above changes would increase the power of the test statistic and allow for easier calculations. 2. Other investigations may subject additional form classes to the four task conditions or extend the number of task conditions investigated. A change in test stimuli might include adjectives and verbs. Additional task condi- tions may include antonyms and homonyms. 115 3. There should be a thorough analysis of the primary and common responses elicited as a function of the race of the subject. Differences in primary and common responses elicited by Black and Latino populations would have profound implications for the standardized testing procedures and clinical rehabilitative techniques currently being used to assess and habilitate aphasic patients. 4. Investigation of restricted word association responses over a greater age range is desirable. A comparison of other age groups ( e.g., young and middle—aged adults) with nonpathological elderly groups of subjects would provide information on changes in response agreement and response diversity across age groups. 5. A comparative study of elderly male and female aphasic patients is important to see if the female aphasic patients have more response agreement than male aphasic patients as was evidenced in the nonpathological sex groups. 6. The effect of years of formal schooling on response agreement and response variability needs further exploration. Closely related to the educational process is the resulting occupational jargon that influences word association responses. The use of occupational jargon as cues for eliciting target words from aphasic patients should be investigated. Bibliography BIBLIOGRAPHY Arenberg, D. "Concept Problem Solving in Young and Old Adults," Journal of Gerontology, 23(1968), pp. 269-282. Balinsky, B. "An Analysis of the Mental Factors of Various Age Groups from Nine to Sixty," Genetic PsychOIOgy Monographs, 23(1941), pp. 191-234. Baltes, P. B. "Life Span Models of Psychological Aging: A White Elephant," Gerontologist, 13(1973), pp. 457-512. Baltes, P. B. and Labouvie, G. V. "Adult DeveIOpment of Intellectual Performance: Description, Explanation and Modification," in C. Eisdorfer and M. P. Lawton (Editors), Psychology of Adult Development and Aging, American Psychology Association, Washington, D.C., (1973» pp. 157—219. Baltes, P. B. and Willis, S. L. "Toward Psychological Theories of Aging and Development," in J. E. Birren and K. W. Schaie (Editors), Handbook of Psychology and Aging, Reinhold—Van Nostrand, New York, 1976. Bayley, N. and Oden, M. H. "The Maintenance of Intellectual Ability in Gifted Adults," Journal of Gerontology, 10(1955), pp. 91—101. Becher, B. A. "A Cross-sectional and Longitudinal Study of the Affect of Education on Free Association Responses,‘ Journal of Genetic Psychology, 10(1960), pp. 23-28. Berman, M. and Peele, L. Mc. Self Generated Cues, A Method for Aiding Aphasic and Apractic Patients, no date. Birren, J. E. The Psychology of Aging, Prentice-Hall, New Jersey, 1964. Botwinick, J. Cognitive Processes in Maturity and Old Age, Springer Company, New York, 1967. 116 117 Botwinick, J. Aging and Behavior, Springer Company, New York, 1972. Buckman, G. F., Jr. "Speech Rehabilitation for the Geriatric Patient," Journal of the American Geriatrics Society, 19(1971), pp. 996-999. Bureau of the Census, Social and Economic Characteristics of the Older Population: 1974, Current Population Reports, Special Studies Series P—23, No. 57, U.S. Department of Commerce, 1975, p. l. Busse, E.W. and Pfeiffer, E. Behavior and Adaptation in Late Life, Little, Brown and Company, Boston, 1969, pp. 1-9. Cattell, R. B. "Theory of Fluid and Crystallized Intelligence: A Critical Experiment," Journal of Educational PsychologY, 54(1963), PP. 1-22. Cattell, R. B. The Scientific Analysis of Personality, Penguin Books, Baltimore, 1965. Chomsky, N. Syntactic Structures, The Hague: Mouton, 1957. Corso, J. F. "Sensory Processes and Age Effects in Normal Adults," Journal of Gerontology, 26(1971), pp. 90—105. Cramer, P. Word Association, Academic Press, New York, 1968. Deese, J. "Form Class and the Determinants of Association," Journal of Verbal Learning and Verbal Behavior, 1(1962), pp. 70-84. Denny, N. W. "Classification Abilities in the Elderly," Journal of Gerontology, 29(1974), pp. 309-314. Doorken, H. "Frequency of Common Associations," Psycho— logical Reports, 2(1956), pp. 407-408. Eisdorfer, C. "Intellectual and Cognitive Changes in the Aged," in E. Busse and E. Pfeiffer (Editors), Behavior and Adaptation in Late Life, Little, Brown and Company, Boston, 1969, pp. 237—247. Eisdorfer, C. and Wilkie, F. "Terminal Changes in Intelli- gence," in E. Palmore (Editor), Normal Aging II, Duke University Press, Durham, 1974. Felton, J. S., Perkins, C. C., Lewin, M. A Survey of Medicine and Medical Practice for the Rehabilitation Counselor, U.S. Department of Health, Education and Welfare, Washington, D.C., 1966, pp. 66-67. 118 Fillenbaum, S. and Rapoport, A. Structures in the Subjective Lexicon, Academic Press, New York, 1971. Flavell, J. H. The Developmental Psychology of Jean Piaget, Von Nostrand, Princeton, New Jersey, 1963. Flavell, J. H. "Cognitive Changes in Adulthood," in L. R. Goulet and P. B. Baltes (Editors), Life Span DeveIOpmental Psychology, Academic Press, New York, 1970. Goodglass, H. and Kaplan, E. The Assessment of Aphasia and Related Disorders, (The Boston Diagnostic Examination), Lea and Febiger, Philadelphia, 1972. Goldfarb, A. I. "Memory and Aging," in R. Goldman and M. Rockstein (Editors), The Physiology and Pathology of Human Aging, Academic Press, Inc., New York, 1975, pp Goulet, L. R. and Baltes, P. B. Life Span Developmental Psychology: Research and Theory, Academic Press, New York, 1970. Guidelines for Identification Audiometry, Asha, February, 1975, pp. 94-99. Herr, V. V. "The Loyola Language Study." Journal of Social Psychology, 46(1957): pp. 81-89. Hooper, F. H., Fitzgerald, J., and Papalia, D. "Piagetian Theory and the Aging Process: Extensions and Speculations," Aging and Human Development, 2(1971), pp. 3-20. Horn, J. L. "Organization of Abilities and the DevelOpment of Intelligence," Psychological Review, 75(1968), pp. 242-259. Horn, J. L. "Organization of Data on Life Span DevelOpment of Human Abilities," in L. R. Goulet and P. B. Baltes (Editors), Life Span Developmental Psychology, Academic Press, New York, 1970. Howes, D. "Application of the Word Frequency Concept to Aphasia," in A. V. S. de Reuck and M. O'Connor (Editors),Disorders of Language, CIBA Foundation Symposium, Little, Brown and Co., Boston, 1965, pp. 47-75. Hutchinson, J. M. and Beasley, D. S. "Speech and Language Functioning Among the Aging," in H. J. Oyer and E. J. Oyer (Editors), Aging and Communication, University Park Press, Baltimore, 1976, pp. 155-174. 119 Human Communication:' The Public Health Aspects of Hearing, Language and Speech Discrders, Monograph No. 7(1968), National Institute of Neurological Disease and Stroke, Washington, D.C., p. 7. Jarvik, L. F. "Biological Differences in Intellectual Functioning," Vita Humana, 5(1962), pp. 195-203. Jarvik, L. F. and Falek, A. "Intellectual Stability and Survival in the Aged," Journal of Gerontology, 18(1963), pp. 173—176. Kahn, R. L., Goldfarb, A. I., Pollack, M. and Peck, A. "Brief Objective Measures for the Determination of Mental Status in the Aged," American Journal of Psychiatry, 117(1960-61), pp. 326—328. Kallman, F. and Jarvik, L. F. "Individual Differences in Constitution and Genetic Background," in J. E. Birren (Editor), Handbook of Aging and the Individual, University of Chicago Press, Chicago, 1959. Kimmel, D. C. Adulthood and Aging, John Wiley and Sons, Incorporated, New York, 1974. Labouvie—Vief, G. "Toward Optimizing Cognitive Competence in Later Life," Educational Gerontology, 1(1976), pp. 75-92. Labouvie-Vief, G. "Adult Cognitive Development: In Search of Alternative Interpretations," Merrill Palmer Press, 1977, (in press). Lansdell, H. "Effects of Neurosurgery on the Ability to Identify Popular Word Associations," Journal of Abnormal ngchology, 81(1973), pp. 255-258. Lieberman, M. "Observations of Death and Dying," Gerontologist, 6(1966)r pp. 70-72. Lienert, G. A. and Crott, H. W. "Studies on the Factor Structure of Intelligence in Children, Adolescents and Adults," Vita Humana, 7(1964), pp. 147-163. Lyons, J. Introduction to Theoretical Linggistics, Cambridge University Press, Cambridge, 1968. Manney, J. D. Aging in American Society, Institute of Geron— tology, UniVersity of Michigan and Wayne State University, Ann Arbor, 1975. Miles, C. C. and Miles, W. R. "The Correlation of Intelligence Scores and Chronological Age from Early to Late Maturity, American Journal of Psychology, 44(1932), pp. 44-78. 120 Miller, G. A. and Johnson-Laird, P. N. Language and Percep— tion, Belkamp Press, Cambridge, 1976. Miller, R. L. The Linguistic Relativity Principle and Humboldtian Ethnolinguistics, The Hague: Mouton, 1968. Milne, J. S. and Lauder, I. J. "Pure Tone Audiometry in Older People," British Journal of Audiology, 9(1975), pp. 50-58. Neugarten, B. "The Future and the Young Old," The Gerontol- ogist, 15(1975), PP. 4—9. ' Onhmacht, F. W. and Pachee, A. M. "Some Effects of Stimulus Modality on Word Association," Psychological Reports, 30(1972), PP- 624-626. Owens, W. A., Jr. "Age and Mental Abilities: A Longifiudinal Study," Genetic Psychology Monographs, 48(1953), pp. 3-54. Oyer, E. J. "Exchanging Information within the Older Family," in H. J. Oyer and E. J. Oyer (Editors), Aging and Communication, University Park Press, Baltimore, 1976, pp. 43-61. Oyer, E. J. "Summary," in H. J. Oyer and E. J. Oyer (Editors), Aging and Communication, University Park Press, Baltimore, 1976, pp. 289-297. Palermo, D. S. "Word Associations and Children's Verbal Behavior," in L. P. Lipsett and C. C. Spiker (Editors), Advances in Child Development and BehaVior: Volume II, Academic Press, New York, 1963, pp. 31-68. Palermo, D. and Jenkins, J. Word Association Norms: Grade School Throggh College, University of Minnesota Press, Minneapolis, 1965. Palmore, E. (Editor), Normal Aging I, Duke University Press, Durham, 1970. Palmore, E. and Cleveland, W. "Aging, Terminal Decline and Terminal Drop," Journal of Gerontology, 31(1976), pp. 76-81. Reese, H. W. "Verbal Mediation as a Function of Age Level," Psychological Bulletin, 59(1962), pp. 502-509. Reynolds, R. J., Bickley, A.C., Champion, 8. and Dekle, 0. "Effects of Mode of Presentation of Stimulus Materials in Word Association Tasks," Psychological Reports, 28(1971), PP. 211-215. 121 Riege1., K. F. The Michigan Restricted Association Norms, Report No. 3, USPHS Grant MH 07619, Ann Arbor, 1965. Riege]_, K. F. "On the History of Psychological Gerontology," in C. Eisdorfer and P. Lawton (Editors), The Psycholo of Adult Development and Aging, American Psychological Association, Washington, D.C., 1973a. Riegeal, K. F. "Dialectic Operations: The Final Period of Cognitive Development," Human Development, 16(1973b), pp. 346-370. Riegrel, K. F. "Language and Cognition: Some Life Span Developmental Issues," Gerontologist, 13(1973C), pp. 478—482. Riegtel, K. F. "From Traits and Equilibrium Towards Develop- mental Dialectics," in W. Arnold (Editor), EEEEEEEE Symposium on Motivation, University of Nebraska Press, Lincoln, 1975. Riegel, K. F. and Birren, J. E. "Age Differences in Associa— tive Behavior," Journal of Gerontology, 20(1965): pp. 153-170. Riegel, K. F. and Riegel, R. M. "Changes in Associative Behavior During Later Years of Life: A Cross- sectional Analysis," Vita Humana, 7(1964), PP. 1-32. Riegel, K. F. and Riegel, R. M. "Development, Drop and Death," Developmental Psychology, 6(1972), pp. 306-319. Rinnert, C. and Whitaker, H. "Semantic Confusions by Aphasic Patients," Cortex, 9(1973), PP. 56-81. Rockstein, M. "The Biology of Aging in Humans—An Overview," in R. Goldstein and M. Rockstein (EditorS)r 222, Psysiology and Pathology of Human Aging, Academic Press, New York, 1975, pp. 1—7. ROSenzweig, M. R. "Word Association of French Workmen: Comparisons with Associations of French Students and American Workmen and Students," Journal of Verbal Behavior, 3(1964), pp. 57—69. schaie, K. W. "A Field Theory Approach to Age Changes in Cognitive Behavior," Vita Humana, 5(1962), PP- 129'141’ Schaie, K. W. and Gribbin, K. "Adult Development and Aging," Annual Review of Psyghology, 26(1975), pp. 63-96. SChUell, H. and Jenkins, J. "Reduction of Vocabulary in Aphasia," Brain, 84(1961), pp. 243-261. 122 Schuell, H., Jenkins, J. and Jimenez-Pabon, E. Aphasia in Adults, Hoeber, New York, 1964. Sefer, J. W. and Henrikson, E. H. "The Relationship Between Word Association and Grammatical Classes in Aphasia," Journal of Speech and'HearingyResearch, 9(1966), pp. 529-541. Sheridan, J. "Restoring Speech and Language Skills," Geriatrics, (1976), pp. 83-86. Spahr, F. "1971 White House Conference on Aging,", Asha, 13(1971). PP. 14-17. Spinnler, H. and Vignolo, L. "Impaired Recognition of Meaningful Sounds in Aphasia," Cortex, 2(1966), pp. 337-348. Spreen, O. "Psycholinguistic Aspects of Aphasia," Journal of Speech and Hearing Research, 11(1968), pp. 453-466. Spreen O. and Wachal, R. S. "Psycholinguistic Analysis of Aphasic Language: Theoretical Formulations and Procedures," Language and Speech, 16(1973), pp. 130-146. Stallones, R. A., Dyken, M. L., Fang, H. C., Heyman, A., Seltzer, R. and Stamler, J. "Epidemiology for Stroke Facilities Planning," Stroke, 3(1972), pp. 360-371. Taylor, 0. "Free Word Association of Aphasics," (unpublished Doctoral dissertation, University of Michigan, 1966). Tresselt, M. E. and Mayzner, M. S. "The Kent-Rosanoff Word Association Norms as a Function of Age," Psychonomic Science, 1(1964), pp. 65-66. Timmermans, B. and Kumin, L. "Free Word Association of Adult Aphasics," Psychological Rgports, 34(1974), pp. 339—390. Thorndike, E. L. Adult Learning, McMillan Company, New York, 1928. Weschler, D. The Measurement and Appraisal of Adult Intelligence, (4th Edition), Williams and Wilkins, Baltimore, 1958. Wiig, E. and Globus, D. "Aphasic Word Identification as a Function of Logical Relationship and Associative Strength," Journal of Speech and Hearing Research, 14(1971), PP. 195~204. 123 Wiig, E. and Globus, D. Language Disabilities in Children and Adolescents, Merrill Publishing Company, Columbus, 1976, pp. 323—324. Wilson, L. A. and Brass, W. "Brief Assessment of the Mental State in Geriatric Domiciliary Practice. The Use— fulness of the Mental Status Questionnaire," Age and Ageing, 2(1973), pp. 92—101. Yamadori, A. and Albert, M. L. "Word Category Aphasia," Cortex, 9(1973), pp. 112—125. Zivian, M. T. Word Identification as a Function of Semantic Clues and Association Strength, Report No. 12, Department of Psychology, University of Michigan, Ann Arbor, Michigan, 1966. Zurif, E. B., Caranazza, A., Myerson, R., Galvin, J. "Semantic Feature Representations for Normal and Aphasic Language," Brain, 1(1974), pp. 167-187. APPENDICES APPENDIX A QUESTIONS USED ON MENTAL STATUS QUESTIONNAIRE The Mental Status Questionnaire 1. Where are we now? 2. Where is this place located? 3. What is today's date? (day of the month) 4. What month is it? 5. What year is it? 6. How old are you? 7. What is your birthdate? 8. What year were you born? 9. Who is the president of the United States? 10. Who was president before him? Source: Goldfarb, Alvin: "Memory and Aging" in The Physiology and Pathology of Human Aging, R. Goldman and M. Rockstein (Eds.) Academic Press, Inc., New York, 1975. 124 APPENDIX B GENERAL AND INDIVIDUAL TASK INSTRUCTIONS GENERAL INSIBUCIIQNS I AM TRYING TO COLLECT INFORMATION ON THE LANGUAGE USED BY SENIOR CITIZENS WHEN THEY ARE ASKED TO SUBSTITUTE ONE WORD FOR ANOTHER. YOU WILL HEAR AND SEE 25 WORDS IN FOUR DIFFERENT TASKS. AN EXPLANATION OF EACH TASK WILL BE GIVEN JUST BEFORE THE 25 WORDS ARE PRESENTED TO YOU. YOU WILL HAVE A CHANCE TO PRACTICE EACH TASK. USE ONLY ONE WORD WHEN ANSWERING. TRY TO RESPOND AS QUICKLY AS POSSIBLE WITH THE FIRST WORD THAT COMES TO YOUR MIND. WES. DO YOU HAVE ANY QUESTIONS? LET US BEGIN WITH TASK l. '125 126 TASK l - SUPERORDINATE YOU ARE TO GIVE A GENERAL CLASS NAME FOR EACH OF THE 25 THINGS PRESENTED 0N SLIDE. FOR EXAMPLE. THE CLASS NAME FOR A ROSE IS FLOWE . THE CLASS NAMES FOR BASEBALL ARE §EQBI OR GAME. THREE MORE EXAMPLES ARE INCLUDED ON THE SLIDE PROJECTOR. WHEN YOU SEE AND HEAR EACH ITEM. YOU WILL RESPOND WITH A SINGLE WORD AS QUICKLY AS POSSIBLE. REMEMBER. YOUR RESPONSE MUST BE A GENERAL CLASS NAME. YOU CAN ADVANCE THE TAPE AND THE SLIDE PROJECTOR BY PRESSING THE WHITE BUTTON ON THE RED BLOCK. 127. TASK 2 - LOCATION YOU ARE TO NAME THE PLACE WHERE YOU MIGHT FIND EACH OF THE 25 THINGS PRESENTED ON SLIDE. IN OTHER WORDS. NAME THE LOCATION OF EACH ITEM. FOR EXAMPLE, A CAR CAN BE FOUND IN A GARAGE OR ON A SIBEEI. A HAT CAN BE FOUND ON A RACK_OR ON A HEAD. THREE MORE EXAMPLES ARE PRESENTED ON THE SLIDE PROJECTOR. REMEMBER. YOUR RESPONSE MUST BE A SINGLE WORD WHICH NAMES THE LOCATION. 128 TASK 3 - SIMILAR YOU ARE TO GIVE ME A WORD THAT PRETTY MUCH MEANS THE SAME AS THE THINGS PRESENTED ON SLIDE. IN OTHER WORDS. NAME A WORD THAT ESSENTIALLY HAS THE SAME MEANING. FOR EXAMPLE; A CUP IS SIMILAR TO A QLASS OR A MU_. A HAT IS SIMILAR TO A CAR OR HELMET. THREE MORE EXAMPLES ARE PRESENTED ON THE SLIDE PROJECTOR. REMEMBER, YOUR RESPONSE MUST BE A SINGLE WORD WHICH ESSENTIALLY MEANS THE SAME. 129 TASK Ll - PART YOU ARE TO NAME A WORD THAT REPRESENTS ONLY A PART OF THE THING PRESENTED ON SLIDE. FOR EXAMPLE. A TOE IS A PART OF THE FOOT. AN ERASER IS PART OF A PENCIL. THREE MORE EXAMPLES ARE PRESENTED ON THE SLIDE PROJECTOR. REMEMBER. YOUR RESPONSE MUST BE A SINGLE WORD WHICH REPRESENTS A PART OF EACH THING PRESENTED. APPENDIX C SAMPLE OF LETTERING USED TO CREATE STIMULI SLIDES 39:3. mEmpH msassflum amoexmq mom womb msflumupmq 130 APPENDIX D ANSWER FORM SHOWING ORDER OF STIMULI PRESENTATION ON THE FOUR EXPERIMENTAL TASKS ANSWER FORM RESTRICTED WORD ASSOCIATION TASKS PRESENTED TO ELDERLY SUBJECTS Name: Birthdate: Chronological Age: yrs. mos. Sex: Educational Level: HS HS Highest Grade Completed: PRE-EXPERIMENTAL DATA MSQ Score: Visual/Reading Screening: Pass _____' Fail Hearing Screening(HTL): dB levels: 500 Hz:____ 1KHz:____ 2KHz :_ 4KHz :_ Residence Elsewhere: ___;No Yes Place (SENERAL COMMENTS: Occupation (s) : Other: 131 132 Subject Mental Status Questionnaire 1. Where are we now? 2. Where is this place located? 3. What is today's date? 4. What month is it? 5. What year is it? 6. How old are you? 7. What is your birthdate? 8. What year were you born? 9. Who is the president of the United States? 10. Who was president before him? Score (8 or above indicates passing score) Degree of Dysfunction: none mild TASK 1: eagle table cabbage bread moth candle anger stomach tiger rug tobacco water knife thief sun house nuisic vuool whiskey QJirl memory street City nurse soldier TASK 2: LOCATION nurse tiger stomach street sun rug cabbage wool memory table thief eagle city whiskey water knife tobacco house soldier moth bread girl candle anger music TASK 3: moth rug SIMILAR 134 SUD city street house eagle table thief tiger cabbage girl bread water candle nurse anger knife memory stomach wool music soldier whiskey tobacco TASK 4: PART moth table cabbage whiskey eagle tobacco candle knife water music bread tiger street city nurse Soldier wool stomach sun house memory anger thief girl rug APPENDIX E ALPHABETIZED PRIMARY AND COMMON RESPONSES AND ASSOCIATED FREQUENCY DISTRIBUTION FOR SUPERORDINATION TASK GNH ems mmm e a me a as a mam m mm N m mm m m e mm m mmmmm m moanmummo> pmmnB sflmum mpoom snon QUSOU wumxmn genus #swEmHmmEmp “somamu moms @008 some mum: :oflumupmssw mcHmem sOHpoEo sofluflwommflp omsommwm soEEoo coapmsflpsosmmsm mo Mmma sow noncommwm QOEEOU mmmnnmo .m woman .N mesm .H aasaaum Hmoaxmq 135 136 maa fleece mNH hmvmvmmoo r-l NOH ml awash N Q mZ mmmaafl> ms3ou soapmasmom momam mamomm coeumNHsmmso muHHMQAOHGSE mflaomouuma cosumooH hnussoo wuecsEEoo MOHM Nm3 magmas Oncommmm GOEEOU maomm .m seed .m mapsmo .v flHsaflum amoexmq 137 mHH NHH wNH Hmuos vm mm ma ma mm ma om NH v-I mfi‘FKDKOONLOO [\ ml somewB Hflmswws maoou mumssm>aem unmasnumsfl msmauso Houuso kuamnm ooswpflmmu msfl>HH um>Ha mmfion mandamze seaweeds oponm soEoB xmm oedema “sowswm mpma amass Hoodoo mmuflsHsHEmw “msflsflfiwm “wawfimm negro omQOMmmm soEEoo Omens .m mmson .m Heed .s .3255 Hmoflmmq 138 mHH Hmuoe mm wN H N N Mr-ImmvNNLONKOOQ‘NNCOON I-‘l ml OOmGH Mu. . man Dragons msflxsflsu “Msflsp wmswm Psoscflosws HOUGHEOH mosmHQEmEmH “mcfluwnfiwfion “HOQEOEOH Hamoms ummm moses mmpmasosx mocmmflaaousfl sOHaOGSM HumuomMOM “uomMOM huasoom setup speaanm Uncommom sosaoo rues .HH huoEwE .OH 3:53 House 139 oeoa HMPOB Ha om ma em wH ml N D mz msou mpcsom msflmsflm “msflm senses “smadupmse usmEsHMpnmusm unmahoflsm “so usmammDEm Omsommmm sOEEOU Dense .NH sausaem Hmoexma 140 MHH sauce u-I meMNNMONNMQVmNMNO N H r-I ml QNMZ smEoz sMHoHsnomu ooa>uom stoflmmowoum “soflmmmwoum sownwm soasmmsooo awesome smflumuflsmesn Hmueamoe DHDQHOS spasms HHHm meEom umumo usmoswuum usmsmemmm oped omsommmm soasoo omsss .MH fissflm Hmofinoq 141 baa mHH Hence NH mm mm mm Hm mm QNMZ HOHHHMB cmuosw> sm5w0fl>umm consumposm “sOPODPOHm Hmsofimmomosd “scammomosm semsom sumaaaas cmE mama H0 dampness smpsmmop wags msflcmAGHSM “msdeflsudm msflsooam “Hooam msaso>oo uooam “msasm>oo “Ho>oo o . . msfluomumo “#mmsmo OmGOMm 0m AHOEOU amaeaom .ma mas .vH 325% House 142 mHH NHH Hmpoe \OQ‘V‘N MI—It-‘l NNNNQ‘MV‘NV‘LOM \OONI—l m r-IONKOO t—lr—lNI—l ml soapmunommsmuu mummsmsouonu xamzmpflm ohmspmou spoon woman #smEo>mm woBsemm unumm mm3mmmmmmm “mommmmm mmszmfln soapomuflp mpflo moBSQ puw>mason Ossm>o mmouppm msmmno mmsflemmesfl mmpflmsfl soaumomflp moon hfioumcm smaocnm Omsommwm soEEoo ummuum .sa SOMEOUM . 0H HHDEHum amoexmq 143 wHH MHH mHH sauce mHH .MHH Nr-IMLDNFKDNKO MQ'KDV'LfiNmm v—l Nr—I ml Hmfiaso w 0 mmusvHGMSM NmZ Q smsumm3 mmum>flss noun mnmndm HCHOM How Usenm mvHHHmumm umsmHm magmas msm>mm£ use: mmuosm pswEmam mmmspamfisn “usmflun moon mmsommwm GOEEOU amass .om magma .mH sum .ma HHSEHum Hmowxmq h: 144 «AH mHH Hmpoe ma mm 0N HH Hm ma ml COOS mmsflonm “oxofim #smHm doommwa “mama page: mSHU CHMHHH> mash swamoum Honpssoom mamanos mamoom “sownmm BMHDSO smE Moose Hmsflfifiso Hmamusn oncommmm sOEEOU ooomnou .Nm wmflnu .HN AHssAum Hmonxmg 145 mHH vHH mHH Hmuos 5H mm ma mm ma ma we NH mm mm ma we ml sum» nuEum3 mwwnm oamflsmums swnflm oestw msfiguoao “moauoao suoao Hosgfla except muoon mmmsm>wn Hocooam umB GHMH canvas pagan usoEwHw mmeAxeAAe “scene mmsommwm soEEoo H83 .mm seeders .em Hwflw3 .MN Assseum Hmoexmg 146 .mmmsommou oeuouocwwoflpfl msflms psonpfls pwnomou on nos oasoo Ho>ma coasmufluo ucmosom mm map muons mommo n o .pmsHEprop mos soaemasmom was mo ANHH u sv usmosog mm How momsommos soafioo map cons poussoo uos mums nonsmowfi wmsommms 0: may .um>030m .mOHosmsgwsw Uncommon sossoo map mo mosam> UHHUESG map mo moses map sflnpflz Haom ouoom can muons pouncepsfl mum Ammzv mswzmsm Uncommon 0: mo Honsds one M Q .pwamfimm coflpoasmom wasmpaw map mp cm>Hm noncommws wumafium one mum poxsmfi mpuoz u w "mmpoz APPENDIX F ALPHABETIZED PRIMARY AND COMMON RESPONSES AND ASSOCIATED FREQUENCY DISTRIBUTION FOR SIMILARITY TASK SHH Hahos Nmm ml apoHB OUCOHOH> HDQEmw Essusms woos some suaaanmuaana one posed: “out: mausong soflvmuemsnw Damage snow coeuoao pwwoscm OmsOQmom cOEEoo weflsmafleflm mo Mmma How noncommmm soEEoo Hogan .H HHDEHum amoexmq 147 .o vHH baa Hmuoe 0H mv mm ma ma OH we mN 0H MN mm ma MN xmz Momma Hermes campsma QEMH QHDQ 3oam usmHMHODMm meonwppds QOSDDOH OHMM HUBOHMHHDCU M flaooooun unmos maaos QHMMSE moxoo mssn mpflsomea oncommwm GOEEOO mapsmo .v momnnwo .m Umoun .N HHDEHum HMOmeA . .. _ ..1 1;... J I. . {Ti .1 r .n.............. ...... .. .Il Eli... Hun-“l":- I II I... .‘n I . I...umhill. .. . . '..|.l .ll . . . IIIAIII 149 baa vHH wad Hobos vN vN mmFQDr—ILOQ‘ON' N Ln wsmEo3 Hmnuofi mmflz sopHmE wflmmma mpma Hmm OHMEOM UHHno GHQOH zoomed H30 Msmn sooamm 30Ho puomnsn manna wmmHHfl> CBC m p Uncommwm sOEEoo :3 .s Davao .m muao .m HHSEH¥W HMUHXOH 150 baa HMDOB N v-I M¢CONONNMOPONNLON v—IN ml smaflmsp usou pwgm #53 DEC: momsmm manaadze OHHUHEOC gosszo mmmepoo sflnmo Bonmssn mason Pcmfienmmo moons wwcommmm GOEEOU Undo: .w HHDEfism amoexmq 151 vHH HMHOB N :—I FMNMO‘MNNNCDNI—ILDFNF N MI mpuozm Hommm wasummm mHOmmHom 3mm “whom Monu Moem wuwsoss usmanmEfl MHOM Mommmp soupso sm>mwao Homage Bowen wmsOMmom coEEoo means .a HHDEHum HOUflxmq *- 152 had Hmpoe N mmMI—INIDNMMFMM I-‘ILn bN OH NN MI m mosouQEoEmH EH03 nomon ouflsgmoE HOHHHE “comma sen Hmaafimumuoo emanamupsn man waummn won usm psmsosu meaxcaru racers #smEchums mosOUMHsHEmH umpsflsmu “msfluwnEmEmu “HOQEmEmH msflaaoomu “maamomu Uses mmmcHDMDOmsow “powwow Sumac gamma mmsommmm sofifiou SNOE .HH whoEmE .oa Hasfiflum amoexoq 153 Hmuoe NN NFNNNQ‘NQ‘NNNLDCDNNONLONNNN ml woeo> sHHoH> many msou UGDOm mmsflmcflm “msem OHCMH poxomu osmflm manusflmm muummnouo muos mmflos MUonE namesuumsfl msoenms mmmsflmmms ucmfiswwpumpsm woemmmHo pang pus omsommmm sOEEOU Dense .NH Aasseum Hmoexmg baa baa Hence \OkaOl‘NKDNNNKDNQ‘ ml use EDOHOQHH msfium>oo mmsflpmmsmo “ummumo N Q m2 soflofisnomp wanmpuo umsuoe OMH3UHE pews nomads anew ouoeoop usmpsmuum ascensmmm open Omsommmm cofiaou 05H .wa onus: .ma HHDEHum HMOAxOA Hence fi' ONQ‘OQ‘KOOAQ'LOMLONMQ t—I ml HOHHHOS pcmmmsmm MHOHHmm uoeomuosd meo>flnm swEOOHHOQ “mosaom h>ms ocHHmE cos pumsm Housmflm umpoo mans oncommmm sosaou seaweed .ma HHSEHPm HDUHNOA 156 mHH sauce NNWNNVNLOGDMMM CD Ln'Q‘NV‘O H N H ml samzueem some“ mmaamfls pnm>wason mssw>m mmaam QNmZ NEED» Moon nosom smmno mmcsH HO>HH mmsflummusfl pumms Hmumomflp “coauwmmep mono Hocflmusoo Hmzon smppman Basses mom swEOUQm mmsomwmm soEEoU pmmuum .na nomsoum .oa Haasaum annexes D NHH does Hmuoe we ma HN N H m N NMNNLDNMNMI‘F N om ma ml CEOHH pummoma Poo N D m2 Hooum possum madam Hmummpmm xmmp pumonmso Hmpssoo umwmsn xon nosmn son scum COOS M Dames Dem: psoeun mwsommmm soEEoo smear .om magma .ma can .ma HHSEHPm amoexmq 158 NHH bHH Hmuoe NH NH NH Hv ml pmmB mmssm wonm mso9nHHmE mm>mmH “HMOH Esm wounmummHo HomHo N D m2 HmHmmum mumnnou HmsHOHssm HMHH umxomfls xoono HmsHEHHo sonHsn ESQ Oncommmm soEEoo ooomnou .NN mmHnu .HN HHDEHum HMUonq 159 MHH Hmuoe N N MQ'LOV‘LGMLDLOKDOMr—Immhm ml OcHB OOp CHOH mom mxHHE HongH canvas OpmsoEOH OOHDm OOH ONE Began adage Omeoo HOOQ HonooHO OmsommOm soEEoo HODOB .mm HHSEHum HOOHqu 160 mHH euros N N mowxokomowq'on I—IN ml OGHB oxpo> schoom HongH OEHm stHp mpsonn ONoon HOOQ HQQOOHO OmsommOm soEEOU smamaez .em AHaEAUm amoexmq 161 .mOmcommOH UHuOHUENOOHpH mchs DUOQHHB pOnomOu On no: pHsoo HO>OH soHquHuo usOonOm mm ng OHOSB mOmmU u .UOCHEHOuOU mm3 ECHpmHsmom One m0 ANHH n sv usOosOm mm sow mOmsommOH soEEoo Onu sOnz pOussoo nos OHOB mOHDmmOE Oms0dmOu 0s Osu .HO>O30m .mOHosOstHm OmsommOH GOEEoo On» m0 mOsHmw OHHOESE Osu m0 Omsou Osu anuHs HHOM Ouoom Osu OsOQB pOuOOHpsH Ono A mzv mHOBOEO OmsomwOH 0: m0 HOQEss One n .UOHmEmm soHsmHsmom NHHOCHO One kn sO>Hm mOmsommOH NHOEHHQ Onu Ono UOMHOE mpuoz H "mOuoz snow I—I dOOSm COHNE HOHHOuOE sOsHH HHOE Adm OOOOHM HOssmHm Adana fi‘ msouwoo maneoHo QDOHO uOMsmHn H003 .mm MHmmfi‘fl‘t—IMMV‘CDN u—l (Or-I r-l OmsodmOm EOEEoo HHDEHum HOOHEOH ml Heron APPENDIX G ALPHABETIZED PRIMARY AND COMMON RESPONSES AND ASSOCIATED FREQUENCY DISTRIBUTION FOR LOCATION TASK mHH Emacs V‘N MMLfiNNNKOV‘PMNOMNLn ml MHOm Eoou HOHHmsv OHQOOQ “EOmHOm OUEHE EOE OEos uHOOn COOS Osman mp30s0 _DH500 EHOHQ Noon usOEDmHm OmGOQOOm COEEOU COHumooq m0 Emma How mOmsommOm c0EEoo sOmsm .H HHDEHum HOUHxOA 162 163 NHH vHH NHH HOHOB NH NH HH mN wN OOH ml manna Ououm OHusmE Omson OEon mumeaoa sousno MOHumOHpsoo OHQOHOpsmo Osoum Ussonm sOpHmm GHOHM M OHnmu Ououm SOHBpsmm smnouHx Omsos pumonmdo xonmeun xon ONHOMOQ OmsommOm GOEEOU OHpsmo .v Ommnnmo .m woman .N HHsEHum HOOHEOH 164 NHH mHH Hence NH mw HH HH hm NNONMMG) v—l ml moooz OOH“ a m w x chussoE uHm OHHOO QNmZ s30“ OOumum whoopuso “OpHmuso mHH0douaOE mmE psmH wsssoo usussoo muHsdEEoo MOHM OmsomwOm EOEEOO OHmmO .o seao .m HHSEHum HOOHxOH 165 mHH Hmuos mH mN Hm Nm mm mN mN ow NH hm ml OHQOD pOxoom OGOQUDHM HOBOHU szou OUOOuum poH psmH EOHUOUESOM Nspssoo muHo pOOuum oHoosom OOHmmo Omsos OEOE sesame Oosmp OoOHHoo mmgommmm £05800 means .m Om50£ .m Hues .s AHsEADm Hmoaxmq 166 NHH hHH Hmhos OH OH om mN HH Nm mN mN msOHOOB “H003 may Omson mmanvoHo “mOnuoHo uOmoHo EHO mpsHE sews CHOHQ OmsommOm GOEEOU Enos .HH >H0EOE .OH Asssaum Hmoaxmq 167 mHH HMH MHH Hmpos mH mOH HMH LOCOV‘MCDNLOMFLDO u—I r-I O u-I \DMLDLOQ‘V‘ r-‘I ml Omson OHOOHM dampemmos Osprnu GOHmH>OHOH OHOpm CHOUOH OHCOM OQOHQ osmegoso OHOQO usOEdsumsH Omsos OEon HHOQ NHOpO>HOms00 quosoo mnousno camp EDHHOHHCSO MHO OmsommOm soEEoo was .wH QWHSG .MH OHODE .NH Hassepm HmoAxdg 168 hHH mHH MHH mHH Hence mH Ho om HH hNH HH mm ml EOOH mcmropax Omson OEon whxw £30“ msoopeso “OpHmuso waussoo ONEHO OHQOOQ “sowsOm sass: HO309 ONUOQ has mxomusmn NEHO m OmsommOm EOEEOO magma .ma DOOuum .sa gooEoww .mH HOHUHOO .mH HHDEHDm HOOHxOH 169 «HH NHH MHH Hmuoa mN mH mN mH mH hN NH mm HH VN mH we mH HH Osoum QUSOQ OOQHQ nusos pHOHm OpuOHmmHo pOOHum OHOpm COmHHm Hagan Omson OEon Ommsmm acme 00M mpooz OOHmssfl umOuom OOHHm< OmGOQOOm COEEOO OUOOQOD .NN aware .HN HOme .om HHDEme HOOHxOH 170 mHH hHH Hmaoe Hm OH mm MN mH mH MN mm Ououm coonm suwaaepmfle NHOBOHQ mmspuon HO>HH OQHQ GOOoo oOMOH OE0£ Undoum DOODOM OmsommOm EOEEOU smxmaez .HN HOHOB .MN HHSEHum HOOHqu 171 .mOmc0dmOH OHOOHOQNOOHOH msHms usosuHB UOQOOOH On #0: pHsoo HO>OH cOHHOpHso usOonOm mm One OHOEB mOmOo u 0 .pOsHEHOHOp mOB soHpOHOQOQ Onu m0 ANHH u av psOosOd mm sow mOmsommOs soEEoo Osu sOss UOucsoo uos OHOB mOsstOE OwsommOs on OS“ .HO>O30m .mOHOQOOgOsm OmsommOH COEEoo Osu wo mOsHO> OHHOEOQ One wo OmsOH One sHsuHB HHOm Osoom OED OMOQB pOuOUHpsH OHO Ammzv mHOzch OmsommOH 0c m0 HOnEss One M Q .UOHQEOm soHuOHDQOQ NHHOUHO OED we sO>Hm mOwsommOn NHOEHsm One OMO pOstE mpsoz n O ”mOuoz VHH m sOHOOsm v OHODm em OQOOnm v QEOH e manpoHo “OOSOOHO H003 .mN HOOOB M OOCOQOOM GOEEOU HHOEHum HOOHqu APPENDIX H ALPHABETIZED PRIMARY AND COMMON RESPONSES AND ASSOCIATED FREQUENCY DISTRIBUTION FOR PART TASK OHOH Hmuoe |—{ M NNNNMLOLDO‘Q‘NMNONV‘NNFN ml sonsOu HOQEOH snonnsum soHpoOOs OmOH sOOE OpOE soHuOuHsuH OHH DH: OuOs soHuOsumssw Osman msHHOOm QOHpoEO COHpHmomme umsmme Hospsoo OosONOQEO OmsommOm soEEou #HOm m0 MmOB How mOmsommOm COEEOU sOmsO .H aassaam amoaxmg 172 173 mHH mHH Hmuoe mm aw HOH MH OH mm MN mH OmsommOm ml 0H3 Ox . xOB Dames OEMHM woos momma #HOOE DOO£3 OUHHm Osson nosoo amuso COEEOU OHUEOO .v OmOQQOo .m women .m Hassaum amoaxmg 174 mHH NHH Hmuos vb mH OH NMNNMNMNLOMMNVQ‘LONNM N N ‘H OmmsHB “msH3 uOnpOOm SOHO sewn mpHOB QHSOEBOH HOsou QHSQSO mpOOHum mOuoum wEsHm UOOH coHuOHsmom OOHQOOQ OmHos poonnonngHOs mHHOQOHuOE HONOE mOm50£ "Omson HHOQ uOHHume maeeaean OmsommOm GOEEOO OHmOO .w sued .m HHDEHpm HOOHxOH 175 NHH sauce (*1 LONKONMQ‘MNLDCDKONNKDNMNNNNN n—I ml wuuHMm Omos OOH: OOOH puOOn pOOQ mpsOn OHHOQ qusOm OHSmHm OHOEOM OOOM mOwO mesp mHudo msHSuoHo umOQuoHo pHHno meOHQ moon quOOQ EHM OmcommOm soEEoo Haas .s AHsEAUm Hmoexmg 176 wHH mHH Hence mH we Hw HOOpm meager mmeman p003 msopsflz OEoou woos nosom sOsouHM OEon ocflsmflsssw “Ousuflsusm sesame Hoop mOsEHno OmsommOm sosEou means .a Omson .w HHSEHDm HOUHxOH 177 mHH OHHH Hmeoe Hm OH 0H ml OmmsHB nusoE mmOH mHOHOOm anemone meAxeAeD "adage OosOHQEOEOH “msHMOnEOEOs “HOQEOEOH HHOOOH umOm OpsHE OmOEH HamuOmHOM “DOmnom soHpOHusOosoo sHOHQ OmsommOm GOEEOU secs .HH NHOEOE .OH Hasseum amoexmg 178 mHH sauce N NNNNNLDNNMNV‘NNNLDCDLONNN Osou mmsHsum UCOOO msHmsHm “msHm HOOEm Osoom OHOom Eguhsu OOOHQ OmOuos OmHos NUOHOE mwa OGOEsspmsH NEOEHOQ psOEsHOusOwsO msHosOp «OocOp mmOn sewn Omsomwwm soEEOU OHmsE .NH Hassaum Hmoaxmg 179 mHH Hmpoe Nr-IC‘ONMNUUNKONNNV'FN ml Ln N GOEOB OEHOMHQS HOuOEoEHng muHHOsomHOQ EOmMOm OosOHuOm OEHOHUOE HOQHOQ mpsOn szom OHOEOm soHpOUSOO sneeDoHo umfioao mOo WEHM OncommOm soEEoo Omsss .MH HHDEHUW HMOHXOH 180 mHH Hmuoe fi' ,4 NNMLONNNLDLOMLOV‘V‘NKONQ'MOLO ml CHOW OH003 msH>OO3 AO>OO3 mqu mow UOOMQH mOH squuOm mOs HOHHOuOE HOE OmsHMm HOQHM UHHQOM smHmOp mcHHO>00 HoHoo DOQHOO HOUHOQ OmOQ OmsommOm soEEoo mas .vH eassaum Hmoaxmq 181 mHH Hmuoe Om MOQEOE OEOs magmas mmEOH COHOOOmuOusH mOmson ossosm EOHMOOHHO nuso OpOnocoo #cOEOo pHOQdmO NOHHO OmsommOm soEEoo uwwupm .sa Haggard Hmoaxmg 184 HNH HNH amaoe (fir-ICC” V‘N mmmmoohmm r—l 00H OmsommOm MI aamma aama a m OmOmH a stm szd “BOA aasoE mmOH UOOE adm uOOm “poow m3OH0 Noon do“ OmmOH coEEOU amaaa .om magma .ma HHDEHum HOOHxOH 185 mHH ooaa amaoa 0H mm OH HN NNV‘Q'NMMCOQ‘Q‘ NN ml EOum OEOEm OGHOOOHG mamma mHooa msHHOOam waOnnoa cowaOd meE cOE aooH mmOH pOOs OmpsOn s50 Mooao HOuOOaOso OmsommOm soEEoo 000On0a .NN MOHEP .HN HHDEHum HOOHxOH 186 NHH amaoe N H MMNNNKOMNLONLOLONNNr-I HM ml DOB mO>O3 ODwOD EOODm sHOa sOONNO OHODOHOE HOaOsHE meaaaaa GOmOprn ONE OpHaosHD UHSHM moat Msaap maaaaa Omflommwm GOEEOU HODOB .MN HHDEHDm HOOHxOH 187 amaos ONMMV‘FMOMMCDM r-Ir—l fi‘ wl DOO£3 HODOB HHOEm Donn Oha DHOE Oaaaaa GHOaw saoo OHDDOQ OHOSOOHO OmsommOm soEEoo smamaas .SN HHSEHDm HOUHxOH 188 Oeoa amaoa mH NH MNMNNNr-‘ILONFUDN N r-l ml QNmz EEO» QDEDOB mpOOanD mosOaDm sHOxm QOOcm HHO dOs cHHosOH QEOH mmOsanoDa OmOOam OUOOHD OHOQHw oaaDOD QDOH0 OmsommOm soEEoo H003 .mN HHDEHDm HOOHEOH 189 .mOmE0mmOa OHDanEmmoHUH mEHmE DEOEDH3 EOEUOOD OE DOE pHsoo HO>OH EoHaODHao DEOoaOm mm OED OaOE3 mOmOU .UOEHEHODOO mO3 EOHDOHEmom OED D0 ANHH n EV DEOOHOQ mm DOD mOmEommOa E0EE00 OED EOE3 UODEsoo D0E OaO3 mOasmOOE OmEommOa 0E OED .HO>O30E .mOHOEOEgOaD OmEommOa EoEEoo OED D0 mOEHO> UHHOEEE OED D0 OmEOa OED EHEDH3 HHOD anom OED OaOE3 UODOOHUEH OaO Ammzv waO3mEO OmEOQmOH 0E D0 HOEEOE OER .UOHQEOO EOHDOHsmom mHaOpHO OED NE EO>Hm mOmEommOa NHOEHHQ OED OaO EOEHOE mpaoz "mODoz I'- llllllllllllllllllllllllllllllllllEs 621431 93 030 Mllllllllllll 312