You are on page 1of 3

1 st TEST ROTTER INCOMPELETE SENTENCEBLANK (RISB) INTRODUCTION PURPOSE OF RISBThe Rotters incomplete sentence blank is an attempt to standardize the

sentence completionmethod for the use at college level. Fort y stems are completed by the subject. These completionsare then scored by compar ing them against typical items in empirically derived scoring manualsfor men and women and by assigning to each response a scale value from 0to6. The total scor e isan index of maladjustmentTHE SENTENCE COMPLETION METHOD The sentence compl etion method of studying personality is a semi structured projectivetechnique in which the subject is asked to finish a sentence for which the first word or wor ds aresupplied. As in other projective devices, it is assumed that the subject r eflects his own wishes,desires, fears and attitudes in the sentences he makes.Hi storically, the incomplete sentence method is related most closely to the word a ssociationtest. In some test incomplete sentences tests only a single word or br ief response is called for; themajor differences appears to be in the length of the stimulus. In the sentence completion tests,tendencies to block and to twist the meaning of the stimulus words appear and the responses may be categorized in a somewhat similar fashion to the word association method. DEVELOPMENT OF ISBTh e Incomplete Sentence Blank consists of forty items revised from a form used by Rotter andWillermann (11) in the army. This form was, in turn, a revision of bla nks used by Shor (15), Hutt(5), and Holzberg (4) at the Mason General Hospital.I n the development of the ISB, two objectives were kept in mind. One aim was to p rovide atechnique which could be used objectively for screening and experimental purposes. It was feltthat this technique should have at least some of the advan tages of projective methods, and also be economical from the point of view of ad ministration and scoring.A second goal was to obtain information of rather speci fic diagnostic value for treatment purposes. The Incomplete Sentence Blank can be used, of course, for general interpretation with a varietyof subjects in much the same manner that a clinician trained in d ynamic psychology uses any projective material. However, a feature of ISB is tha t one can derive a single over-all adjustmentscore. This over-all adjustment sco re is of particular value for screening purposes with collegestudents and in exp erimental studies. The ISB has also been used in a vocational guidance center to select students requiring broader counseling than was usually given, in experim ental studies of the effect of psychotherapy and in investigations of the relati onship of adjustment to a variety of variables. PSYCHOMETRIC PROPERTIES 1.RELIA BILITY Since the items on an incomplete sentence blank are not equivalent, the o dd even technique for determining reliability is not applicable and would tend t o give minimum estimate of internalconsistency. Therefore items on the ISB were divided into two halves deemed as nearlyequivalent as possible. This yielded a c orrected split-half reliability of .84 when based on therecords of 124 male coll ege students, and .83 when based on 71 female students.Inter-scorer reliability for two scorer trained by the authors was .91 when based on malerecords and .96 for female records.2.VALIDITY The Incomplete Sentence Blank was validated on gro ups of subjects which did not include anyof cases used in developing the scoring principles and the scoring manuals. Scoring of the blankswas done blindly the sco rer never knew whether the test blank was supposed to be that of amaladjusted or an adjusted subject.Validity data were obtained for the two sexes separately si nce the scoring manuals differ. Thesubjects include 82 females and 124 males who were classified as either adjusted or maladjustedi.e., as needing personal coun seling or as not needing such counseling.A cutting score of 135 provided a very sufficient separation of adjusted and maladjustedstudents in the data collected above3.NORMS A distribution of scores on the ISB for a representative college freshman popul ation wasobtained by giving the Incomplete Sentences Blank to 299 entering fresh man at OhioState University. A comparison between the median percentile ranks on the Ohio StatePsychological Examination of the sample and of the total freshman population showed adifference of approximately two percentile points. The agree ment between corresponding firstand third quartile points was very close. It was interesting to find that the correlation coefficient between the Ohio State Psy

chological Examination scores and ISB scores for the selectedfreshman sample was only .11. This is in accord with a general feeling that a very littlerelationsh ip would exist between intelligence and scores on the personality measure such a s theIncomplete Sentence Blank ADMINISTRATION Incomplete sentence blank-college formName sex.. Age.. Marital state.. Place. Date.. Complete these sentences to express your real feeling. Tr make complete sentences.1. I like....2. The happiest time The printed instructions are given on the page which statecomplete these sentenc es to express your real feelings. Try to do every one. Be sure to make a complet e sentence.No further instructions are given except to repeat the printed instru ctions if necessary and to urge subjects tocomplete all the items. Administratio n to a group of any number of subjects is possible. Theapproximate average time for administration is twenty minutes. The RISB was administered on a female of 2 0 years of age. The test consisted of fortyitems which had to be completed by th e subject. The test was administered in a class setting.Subject was seated comfo rtably on the chair. The room was well ventilated and well lighted. Theabove ins tructions were given to the subject before he started the test. The subject comp leted thetest in 20 minutes. SCORING THE USE OF SCORING MANUALSentence complet ions are used from examples in the scoring manuals by assigning anumerical weigh t from 0 to 6 for each sentence and totaling the weights to obtain the overallsc ore. The scoring examples in part II of this manual are given to facilitate the assignment of weights to responses. They are from ISB responses of 58 male and 53 female colle ge students,ranging from extremely well adjusted person to those judged to be in need of psychotherapy.Since the scoring examples are illustrative and represent ative of common responses with nointent to list all possible sentence completion s, a set of scoring principles will be presented.These principles are intended t o aid in determing the correct weight for a completion when avery similar statem ent cannot be found in the scoring examples.In order to provide the potential us er of the ISB with supervised experience beforeattempting to score clinical or exp erimental records. The correct scoring for these records isgiven at the end. The se examples will enable the clinician to check his scoring against that of theau thors. They may also use by a clinic supervisor to check the scoring ability of any student or general scorer.Sentence completion is used for illustrative purpo ses in the following discussion are takenalmost entirely from the manual. SCORI NG PRINCIPLES OMISSION RESPONSE Omission responses are designated as those for w hich no answer is given or for which thethought is incomplete. Omissions and fra gments are not scored.It is recognized that in a clinical situations are occasio nally provocative since they may point toareas which the individual does not rec ognize or cannot bring himself to express.For all responses which are subsumed u nder the heading of incomplete thoughts or omissions,no scoring is made. After t he remainder of responses is prorated by the formula {40 / (40-omissions)} times the total scores however, if there are more than 20 omissions, the paper iscons idered unscorable for all practical purposes.For example, Most girls . . . dont appeal to me except sexually because; or I hate. . . thethought of goin g home sinceCONFLICT RESPONSES C or conflict, responses are those indicating an un healthy or maladjusted frame of mind.These include hostility reactions, pessimis m, symptom elicitation, hopelessness and suicidalwishes, statements of unhappy e xperiences, and indications of past maladjustment. Responses range from C1 to C3 according to the severity of the conflict or malad justedexpressed. The numerical weights for the conflict responses areC1=4C2=5C3= 6Typical of the C1 category are responses in which concern is expressed regardin g such thingsas the world state of affairs, financial problems, specific school difficulties, physical complaints,identification with minority groups, and so on . In general it might be said that subsumed under C1 are minor problems which ar e not deep-seated or incapacitating, and more or less specificdifficulties.More serious indications of maladjustment are found in the C2 category. On the whole theresponses refer to broader, more generalized difficulties than are found in C 1. I Included hereare expressions of inferiority feelings, psychosomatic complai

nts, concern over possible failure,generalized school problems, lack of goals, f eeling of inadequacy, concern over vocationalchoice, and difficulty in heterosex ual relationships as well as generalized social difficulty.Expression of severe conflict or indications of maladjustments are rated C3. Among thedifficulties fo und in this area are suicidal wishes, sexual conflicts, severe family problems, fear of insanity, strong negative attitudes toward people in general, feelings o f confusion, expressionof rather bizarre attitudes, and so forth.For example , I like. . . to know if I am crazy. This type of response will lie in C3 category.The happiest time . . .is over and this type of response will lie in C2 category. I want to know. . .about life, this type of response will lie in C1 category POSITI VE RESPONSES P or positive responses are those indicating a healthy or hopeful fra me of mind. These areevidence by humorous or flippant remarks, optimistic respon ses, and acceptance reactions.Responses range from P1to P3 depending on the degr ee of good adjustment expressed in thestatement. The numerical weights for the p ositive responses areP1= 2P2=1P3=0 In the P1 class common responses are those which deal with positive attitudes to ward school,hobbies, sports, expression interest in people, expression of warm f eeling toward someindividual and so on.Generally found under the heading of P2 a re those replies which indicate a generalized positivefeeling toward people, goo d social adjustment, healthy family life, optimism and humor.Clear cut good natu red humor, real optimism, and warm acceptance are types of responses whichare su bsumed under the P3 group. The ISB deviates from the majority of the test in tha t it scoreshumorous responses.For example, I like. . . to have good time, this ty pe of response will lie in P1 category. Thehappiest time . . . is yet to come, this type of response will lie in P2 category. Back home. . .are many friends, this type of response will lie in P3 category. NEUTRAL RESPON SES N or neutral responses are those not falling clearly into either of the above categories. Theyare generally on a simple descriptive level. Two general types o f responses which account for alarge share of those that fall in the neutral cat egory. One group includes those lacking emotionaltone or personal reference. The other group is composed of many responses which are found asoften among maladju sted as among adjusted individual and through clinical judgment could not be leg itimately place in either C or P group. All the N responses are scored 3.For exa mple, Most girls . . . are females or When I was child . . . I spoke as a child. Thesetypes of respo nses will lie in neutral responses. INDEPENDENT SCORING OF ITEMS Each response i s to be scored and evaluated independently of all others, except when it is clea r-cut reference to a previous statement. It is, of course, important in the scor ing of any papers toavoid the halo effect as much as possible so that the measur ement can be reliable. This is equallynecessary here for, if each response is no t scored independently of all others, there is a tendencyto rate all responses i n light of the over-all picture.In some cases a response refers directly to a pr evious item, and it would not be reasonable toscore it independently of the firs t. In such an instance, therefore, a previous response must beused in the evalua tion of the later one.