ADAPTIVE ABILITY TESTS User Manual
Contents
1. RRR 9 7 5 2 3 3 3 2 2 2 2 5 4 2 2 2 1 2 2 3 3 1 MEAN 37 02 SAMPLE SIZE 139 VARIANCE 159 29 STANDARD DEVIATION 12 62 1 April 2004 Page 422. 1 April 2004 Page 34 Selby amp Mills Limited Adaptive Ability Tests B Group Characteristics There are two factors that the user needs to consider with respect to the group from which the selection is being made and the group who are currently in post The selection ratio is the ratio of the number of openings to the number of applicants Whenever this ratio is less than one then the employer can be selective i e There are more applicants than openings The general principle is that the lesser the selection ratio is then the better will be the quality of the selected employees as long as the relationship between the selection test and the criterion of success in the job is a value greater than zero This means that if one hundred applicants apply for a single job then a test of low validity e g 0 1 will still be perfectly valid given that the test is reliable The relationship between test validity and selection ratio to percentage of selected applicants who will be successful can be found in McCormick amp Ilgen Industrial Psychology 7th Edition 1980 p133 The Base Rate is the proportion of individuals who can be regarded as currently successful in the job for which selection is going on For a given validity the fewer people who are currently successful in the job the larger percentage increase in satisfactory applicants will occur with the introduction of a predictor 11 2 Relationship with Other T
3. TABLE 12 Intercorrelation matrix locator items The results are shown in Table 6 Language Numeric amp Administrative modules Graduates N 445 ERR Numeric Module Administrative Module Language module Numeric module TABLE 13 Intercorrelations matrix Administrative module amp Administrative module sections Graduates N 445 LU Addresses Section Codes Section Administrative Module Numbers Section Addresses Section Codes Section TABLE 14 Intercorrelation matrix The three ability test modules amp administrative module sections Language Numeric amp Administrative modules Graduate N 445 Numeric Administrative Numbers Addresses Codes Module Module Section Section Section Language Module Numeric Module Administrative Module Numbers section Addresses section UU dct Sq 1 April 2004 Page 32 Selby amp Mills Limited Adaptive Ability Tests 11 VALIDITY Broadly speaking the validity of a test is an indication of how much a test is measuring what it is supposed to be measuring Again there are a number of different types of validity depending on what aspect the user is looking at The most common forms of validity are as follows A Face Validity This is a measure of how much a test appears to measure what it is supposed to measure Ability tests will generally have a high face validity for example a candidate will clearly understand that a question about percentages wi
4. Mills Limited Adaptive Ability Tests Diagram 3 STENS and the normal curve 2 28 6 68 15 87 30 85 50 00 69 15 84 13 93 32 97 72 Cumulative percentage of scores Norms are provided as part of the software package However if it possible for the user to generate their own norms This facility enables the user for example to generate norms that would then be available as part of the software Such norms could be generated for different job groups for example technical staff or management trainees This gives the user norm groups which are directly relevant hence improving the accuracy of their decision making 1 April 2004 Page 15 Selby amp Mills Limited Adaptive Ability Tests 6 AVAILABLE NORMS This section gives information about the norms that are currently available for the Adaptive Ability Tests These are provided as part of the software and are reproduced here for reference 6 1 Clerical Norms TABLE 1 NORMS FOR OPERATIVES Sample Size 256 Sample Description Employees from a large securities company in UK ABILITY SAMPLE STEN SCORES MEAN S MODULE SIZE TANDARD DEVIATION LANGUAGE 91 126 127 NUMERIC 61 92 93 ADMINISTRATIVE NUMBERS 39 48 49 52 53 56 57 ADDRESSES 30 42 43 49 50 59 60 CODES S 35 41 42 68 69 84 85 OVERALL 106 146 147 174 175 188 189 6 2 Graduate Norm TABLE2 NORMS FOR GRADUATE POPULATION Sample Size 445 Sample Description the s
5. altered to suit this mode of presentation and a simple and pragmatic method of tailoring tests adopted Further data collection is at present ongoing enabling the use of more widely accepted probability model methodology Background The Adaptive Ability Tests were developed in a series of stages From an initial item pool containing items previously tried in the Army Entrance Test project items were selected to represent appropriate difficulty levels and content These items were trialled in pencil and paper format using samples of personnel employed by some of our clients It was this first sample which was used to set up the difficulty levels one two and three on each of the modules A total sample size of 180 people consisted of 95 new entrants 85 middle and senior management volunteers From the initial analyses of these data it became apparent that for the language and numeric modules there may have been a problem discriminating performances at the lower end of the ability range It was for this reason that a further sample of school children in two Bicester schools 1 April 2004 Page 18 Selby amp Mills Limited Adaptive Ability Tests were used to locate the Level 0 Language and Numeric tests A total sample size of 68 consisted of 27 St Mary s Junior School 4th year pupils age 10 11 41 Bicester Community College 1st year pupils age 11 12 The pencil and paper forms of the tests were item analysed using a spl
6. 1 point and the number that were passed to grade A TABLE 17 CORRELATION COEFFICIENTS BETWEEN THE ADAPTIVE ABILITY TESTS AND GCE A LEVEL EXAMINATIONS A Level Point Score Number at Grade A Language Numeric Administrative Numbers Addresses Codes 1 April 2004 Page 37 Selby amp Mills Limited Adaptive Ability Tests Table 18 Pilot Study Trials For Management Supervisory Operative And School Sample Groups On Paper And Pencil Versions NUMERIC LANGUAGE NUMBER ADDRESS SCHOOLS NUMERIC LANGUAGE Sample Management Supervisory Operatives Management Supervisory Operative Management Supervisory Operative Management Supervisory Operative Manager Supervisory Operative Combined Bicester St Mary Combined Bicester St Mar Sample Size Variance Standard Deviation 6 1 3 2 og 11 5 5 6 5 3 5 7 3 8 5 8 7 9 4 7 5 2 9 2 5 9 10 7 8 5 6 2 8 8 4 9 3 9 5 9 1st April 2004 Page 38 Selby amp Mills Limited Adaptive Ability Tests TABLE 19 FREQUENCY DISTRIBUTION COMPUTERISED LANGUAGE MODULE Score Frequency Cumulative Percentile T Score Z Score Sten Score Frequency Cumulative Percentile T Score ZScore Sten Frequency Frequency CQ N Oo 6 7 5 6 4 6 4 3 3 4 4 1 3 5 4 5 4 5 5 2 5 4 3 Ae N WO DBD N N F N N Q MEAN 53 77 SAMPLE SIZE 256 VARIANCE 699 3602 STANDARD DEV
7. Appendix DIAGRAM 5 TYPICAL PROFILES FOR EACH GROUPING LANGUAGE NUMERIC ADMINISTRATIVE NUMBERS ADDRESSES CODES M MANAGERS H HIGH PERFORMANCE GROUP S SUPERVISORY OPERATIVES Table 17 see Appendix gives a summary of the pilot study trials for management supervisory operative and school groups in the paper and pencil versions Tables 18 to 23 see Appendix show frequency tables for each of the test modules from this initial computerised version pilot 1 April 2004 Page 26 Selby amp Mills Limited Adaptive Ability Tests 9 RELIABILITY As with all instruments that are used for measuring tests always include some error of measurement Generally this error will arise from three different sources i The conditions under which the test is taken may vary from one situation to another ii The way the candidate is feeling will vary e g their health iii The contents of the test may favour certain individuals It is due to these inconsistencies that we need to have statistics which indicate the robustness of a particular measure In psychological testing this robustness is known as reliability There are three main types of reliability measure however the underlying assumption is the same the test is measuring an individual s ability in a reliable manner then that individual s scores will be consistent Reliability is expressed as a numeric value between 1 and 1 A correlation of 1 describes perfect co
8. all types of validity including concurrent predictive and content Generally the construct validity of a test is established through a slow laborious process of gathering evidence from many experiments and observations on how the test is functioning Construct validity can be also be assessed by correlations with other tests or measures of the same characteristic Thus one particularl language ability test should correlate with another test of language ability 11 1 Factors Influencing the Utility of a Test As well as reliability and validity there are other factors which help us to assess the usefulness of a selection device These include the appropriateness of the criterion as well as characteristics of the group who are applying for the job and the group who are currently in the job A Criterion The criterion that is used should be relevant an reliable The criterion should be reliable in that it should not be affected over time it should also be able to discriminate appropriately e g between high and low performers The criteria should also be relevant in that it is valid for its purpose This is to say to what extent is the actual criteria that is being used e g sales figures relevant to the ultimate criterion e g success as a sales manager In real terms the test user also needs to consider how acceptable a criteria is how much it will cost to measure how representative it is and how measurable it is among other factors
9. are entirely administered by the computer However it is also important to consider the conditions in which the computer is used The computer should be set up in a quiet room free from interruptions It is also important the contrast brightness and colour of the screen is to the satisfaction of the candidate The screen should also be set to avoid any glare for example from a window or internal lighting Once the candidate has been settled at the screen the session is administered by the computer At the beginning of each of the modules there are a set of screens which introduce that particular module These screens are self paced i e the next screen will not appear until the candidate presses the enter key on the keyboard These initial screens show the candidate examples of the type of question they will encounter and how to select an answer Timing for a particular test does not begin until the first question appears on the screen Each test starts with a short locator test which determines the level at which the applicant will start answering Once allocated to a given level in a subject area the candidate s performance is continuously monitored and they are shifted to different levels if they are performing above or below average from that level In this way the candidate has the maximum number of questions around their particular level of ability Questions are presented sequentially one at a time The candidate will not
10. important factor which is associated with being able to do the job for which you are applying NB If a candidate queries the validity of the test for their ethnic or cultural group reassure them of the integrity and validity of the entire procedure I m a member of a minority group These tests will or should have been validated for use with respect to the application for which they are being used This means that if a test measures a candidate s numeric ability the organisation using it should have determined in advance that this kind of ability is an important requirement for success in the job or career for which the candidate is being considered Therefore irrespective of the candidates ethnic or racial origin their gender or age they need to be able to cope with the type of work requirement which is evaluated by the test or tests 1st April 2004 Page 11 Selby amp Mills Limited Adaptive Ability Tests NB When you are administering to someone and you feel the candidate is genuinely uncomfortable use your discretion to query this with the candidate Generally believe what candidates tell you in this respect If you are unsure whether to intervene and decide not to do so but feel the candidate is genuinely experiencing discomfort due to the test make a note of this on the paper and discuss it with the authorised test user in your organisation or if you are that person please call for advice Always err in favour of the
11. in a single administration C Reduced administrative burden The use of the PC eliminates a lot of the paraphernalia that the traditional paper and pencil users have had to contend with Answer sheets test booklets administration cards scoring cards pencils rubbers and stopwatches are no longer needed Test administration and scoring are carried out entirely automatically The candidate answers each question at the computer Once the candidate has completed a particular module the results are stored on the computer The administrator may then produce a hard copy of the results Scoring is carried out automatically and feedback can be given instantaneously D Developed specifically for use in the UK and Europe The Adaptive Ability Tests were wholly developed within the UK The initial items were drawn from the extensive results of ability tests which have been used in the regular and territorial armies for over a decade The tests have also been trialled in a number of UK companies as well as with an extensive sample of graduates across the UK Lower levels of the numeric ability module were trialled at schools in Oxfordshire Improved reliability The modules are entirely administered by the computer This eliminates the bias associated with human administrations The computer will guarantee 10090 consistency in administration Similarly the results are scored by the computer and again this will be independent of human error
12. individuals it is possible to generate a number which represents the Standard Error of Measurement for this test This will give a range within which the true score of the individual will fall For example If SEm 2 marks then 68 of the time the individual s true score will lie between 4 and 8 i e 6 2 marks As with any normal distribution if we were to allow two SEmv s either side of the score i e 6 4 marks then this is 95 probability that the individual s true score will lie within this range In this example the score of six will be affected by factors such as the individual s true score will lie within this range In this example the score of six will be affected by factors such as the individual s health the nature of the test administration and whether the test items suited the individual Hence the individual s true score independent of these factors will 9596 of the time be between 2 and 10 The use of the SEm is most obvious in borderline cases However it is good practice to recognise that the scores you obtain relate to human beings and at all levels score evidence should be sought from other sources such as tests interviews and so on to take account of this error 1 April 2004 Page 29 Selby amp Mills Limited Adaptive Ability Tests 9 2 Standard Error of Differences In much the same that individual test scores can be differentiated using the Standard Error of Measurement and evaluated to some exten
13. major part of the occupational test market in the UK They have until recently been paper and pencil materials In the last decade the PC has become an integral part of virtually all work environments This has opened up new opportunities in the field of occupational testing and especially down the route of ability testing The Adaptive Ability Tests consist of three modules which are offered to the user in software form and can be loaded onto most PCs The three modules are 1 Language Ability 2 Numeric Ability 3 Administrative Ability The Adaptive Ability Tests were developed to offer the user a number of key benefits over traditional paper and pencil tests A Each of the three test modules will assess the entire range of ability in a single test administration lasting no more than twenty minutes To do this the Adaptive Ability Tests utilise sophisticated hunting algorithms as part of the software design These algorithms ensure that the candidate will need to answer fewer questions to indicate their level of ability to the same accuracy as traditional methods Such tests are called Adaptive Ability Tests The hunting algorithms work as follows The computer continually monitors the speed and accuracy at which the candidate is working This then enables the computer to select only those items which lie close to the candidates ability level If the candidate answers a set of questions correctly in a short space of time the comput
14. 2 GRADUATE NORM fcc tein ue e Ee e e sree EE eden 16 7 QOEM pU 17 8 BACKGROUND AND DEVELOPMENT eee eee eere eee enne eee tn ne etta an sese eo Pese eese eese eaae estate 18 8 1 eese aatem ce 18 8 2 BACKGROUND a E EE E S EE E 18 8 3 MEASUREMETERS 20 8 4 COMPUTERISEP V ERSION 22 8 5 STANDARDISATION OF THE TESTS eee eet eter 25 9 RELIABILITY 27 9 1 STANDARD ERROR MFASUREM PN ETA E RR 28 9 2 STANDARD ERROR OF DIFFERENCEG ccssesssscscsssssssssccsccscsesssssessecsceesssssssssescsssesssesseesesesesssesssevesceesseesees 30 9 3 RELIABILITY OF THE ADAPTIVE ABILITY TESTS ccccccccccsesssssscescccceessssscssecscessessscssessescsesseesesscesseessasees 31 10 INTER RELATIONSHIPS OF THE ADAPTIVE ABILITY TEST 32 11 33 11 1 FACTORS INFLUENCING THE UTILITY OF A TEST eene emen enne teneret eterni tense eren ene 34 11 2 RELATIONSHIP WITH OTHER TESTS 35 11 3 RELATIONSHIP BETWEEN THE ADAPTIVE ABILITY TESTS AND EDUCATIONAL EXAMINATION VIVAS 36 1 April 2004 Page2 Selby amp Mills Limited Adaptive Ability Tests 1 INTRODUCTION Ability tests have been used in the UK over several decades and the ability test market has formed by far the
15. F Automatic data storage As a candidate completes a particular test the results are automatically filed This means that the results remain accessible at all times and can be used as a 1st April 2004 Page 4 Selby amp Mills Limited Adaptive Ability Tests database for producing additional norms They can be archived for future retrieval G Comprehensive Support and Updates As these adaptive tests were developed by us we are able to offer full telephone support and software updates Refinements in the software will automatically be available As part of our development programme the user will be regularly updated with new norms and validity studies to add to their manual and users guide 1 April 2004 Page 5 Selby amp Mills Limited Adaptive Ability Tests 2 THE ADAPTIVE ABILITY TEST MODULES 2 1 Description The Adaptive Ability Tests were developed and computerised for use in the selection training and career development of personnel in the UK The modules are designed for all types of company personnel these include 1 Operatives 2 Clerical Supervisory Staff 3 Management The Test Modules include measures of Language Numerical Ability and Clerical Administrative skills The modules have different levels of difficulty This means that the individual taking the test will only have to answer the questions suited to their level They do not have to answer a large number of redundant questions which would have to be ans
16. IATION 26 44 N N M DBF Q Q e 1st April 2004 Page 39 Selby amp Mills Limited Adaptive Ability Tests TABLE 20 FREQUENCY DISTRIBUTION COMPUTERISED NUMERIC MODULE Score Frequency Cumulative Percentile T Score Z Score Sten Frequen 1 2 1 1 1 1 1 1 1 1 1 2 4 4 1 4 7 4 9 6 6 4 8 m Aae N BP PN BD MEAN 42 42 SAMPLE SIZE 180 VARIANCE349 7 STANDARD DEVIATION 1 April 2004 Page 40 Selby amp Mills Limited Adaptive Ability Tests TABLE 21 FREQUENCY DISTRIBUTION COMPUTERISED ADMINISTRATIVE MODULE OVERALL Score Frequency Cumulative Percentile T Score ZScore Sten Score Frequency Cumulative Percentile T Score Z Score Sten Frequency Frequency RPP RRP RP RP amp UO NO ERU O o MEAN 104 14 SAMPLE SIZE 139 VARIANCE 2058 5588 STANDARD VARIATION 45 37 OQ gt 1 April 2004 Page 41 Selby amp Mills Limited Adaptive Ability Tests TABLE 22 FREQUENCY DISTRIBUTION COMPUTERISED ADMINISTRATIVE MODULE NUMBERS SECTION Score Frequency Cumulative Percentile T Score Z Score Sten Frequency gt
17. RS2 ADDRESS2 CODES2 NUMBERS3 ADDRESS3 NUMERIC 1 3 44 LANGUAGE 2 3 44 3 44 NUMERIC 2 3 46 3 46 3 44 NUMBERS 1 2 75 2 75 2 75 2 78 ADDRESS 1 3 32 3 32 3 92 3 34 2 60 CODES 1 2 80 2 80 2 80 2 73 3 99 2 65 NUMBERS 2 3 08 3 08 3 08 311 2 00 2 95 2 95 ADDRESS 2 3 83 3 83 3 83 3 85 3 22 3 27 3 52 CODES 2 449 449 449 3 78 3 99 4 40 4 03 4 24 4 76 NUMBERS 3 2 86 2 86 2 86 2 90 2 00 2 74 2 07 2 45 3 34 4 08 ADDRESS 3 3 24 3 24 3 24 3 26 2 49 3 11 2 55 2 86 3 65 4 34 2 64 CODES 3 2 69 2 69 2 69 2 72 1 73 2 54 1 81 2 27 3 18 3 95 1 93 2 43 1 April 2004 Page 30 Selby amp Mills Limited Adaptive Ability Tests 9 3 Reliability of the Adaptive Ability Tests Test retest correlations are shown in Table 11 These are preliminary results and it should be noted that the sample size is small On average the period between each test administration for the sample was a fortnight TABLE11 INTERNAL CONSISTENCY OF THE ABILITY TESTS TEST RETEST CORRELATIONS N 16 COMPOSITE GROUP Module Second Administration Numeric Language Administrative First Numeric Administration Language Administration EIL ZI 1 April 2004 Page 31 Selby amp Mills Limited Adaptive Ability Tests 10 INTER RELATIONSHIPS OF THE ADAPTIVE ABILITY TEST MODULES The intercorrelations of the three modules of the Adaptive Ability Tests are shown below Additional correlations are provided for the three sections of the administrative ability module
18. Selby amp Mills Limited Tregoning Mill St Keverne Helston Cornwall Tel 44 0 1326 280382 TR12 6QE Email info selbymills co uk United Kingdom Web site www selbymills co uk Selby amp Mills Limited 2004 Selby amp Mills Limited Adaptive Ability Tests CONTENTS 1 GSINTRODUGCTION ECCE 3 2 THE ADAPTIVE ABILITY TEST MODULES u0u ccssscccssseecssssccccsssecccssccccsssccccssseccsessccccessecccessnees 6 2 1 DESCRIPTION ete ee tette ite E eie 6 2 2 EANGUAGEJZBIEEDYGG d t mec e M LIU LL EA D DI UM LI 6 2 3 tenter Secs ae ca eee he arch 6 24 ADMINISTRATIVE ABILITY cccccesscsscceceesssssccsceccscsesssessescescsesssesscsseeceessesesssescesssssessscscecsesssessseveessessasees F 3 ADMINISTRA TION eve ee aises 9 3 1 2 r ce ive 9 3 2 QUESTIONS AND ANSWERS FOR ADMINISTRATORS ccssccccesssececesscececseeecesssececssseeceeseeecsssseeeenseeeeeenaes 10 4 SCORING THE ADAPTIVE ABILITY TESTS eesseoeessecescsecssoceessececsseosscsecssooeessoceessecessseessooeessoceese 13 5 INTERPRETATION THE USE OF NORMS 4 ee ee eene eee tnos ee tone eese enses eese eese enne 14 6 AVAILABLE NORMS ee epa are Pose Ue io Pe eara ERR 16 6 1 CLERICAL NORMS A ee Edd 16 6
19. ample was drawn from eleven different universities and polytechnics across the UK the majority of the sample was aged between 20 and 22 ABILITY STEN SCORES MEAN STANDARD MODULE DEVIATION LANGUAGE 73 82 83 92 103 112 113 122 NUMERIC 83 90 91 98 108 115 116 123 ADMINISTRATIVE NUMBERS 0 41 42 46 47 51 52 56 57 61 62 66 67 72 73 77 78 82 61 925 ADDRESSES 0 21 22 30 31 40 41 50 51 59 60 69 70 78 79 88 89 98 60 690 CODES 0 27 28 45 46 64 65 83 84 102 103 120 121 139 140 158 159 176 102 797 OVERALL 0 113 114 141 142 168 169 196 197 224 225 252 253 280 281 308 309 335 225 412 1 April 2004 Page 16 Selby amp Mills Limited Adaptive Ability Tests RELIABILITY This chapter is aimed at giving a brief introduction to the reliability of the Adaptive Ability Tests Please note that complete details can be found in the technical section of the manual Reliability and Validity are the two measures which indicate if a test will be effective for the user For a test to be valid it must be reliable Reliability is a measure of the consistency and robustness of a measure Reliabilities are always expressed in terms of a correlation coefficient which will line between 1 and 1 We can consider a coefficient of 65 or above as good and a coefficient of 9 or above as exceptional There are several different methods of assessing the reliability of a test The table below presents the test re test coefficients of the three modules of th
20. ation Attainment This section of the technical manual gives data about the relationship between performance of a group of graduate level candidates and their GCE O level and A level results GCE O Levels The correlations between the Adaptive Ability Test and GCE O levels is given in Table 15 The sample was a group of second and third year undergraduates who had taken their O levels four or five years earlier in most cases The Ability Test modules were correlated with O level points score where a grade A 3 points grade B 2 points and grade C 1 point the grade at which O level maths was passed and the grade at which O level English language was passed 1st April 2004 Page 36 Selby amp Mills Limited Adaptive Ability Tests TABLE 16 CORRELATION COEFFICIENTS BETWEEN THE ADAPTIVE ABILITY TESTS AND GCE LEVEL EXAMINATIONS Module O level English Points Score Language Grade Numeric Language Administrative Numbers Addresses Codes 11 3 2 GCE A levels The correlation between the Adaptive Ability Tests and GCE A levels is given in Table 17 The sample was used This means that the results of the ability tests have been correlated with A level performance taken on average two or three years earlier Results were correlated with the number of A levels points score where grade A 5 points 4 points C 3 points D 2 points and E
21. be presented with the next question until they have answered the question they are on Timing of the test occurs automatically the test will be terminated automatically and this may occur at any stage during a test A screen will inform the candidate that they have completed the test and that they should inform the administrator Section 3 1 below gives an example of how to run the test session and what to say to the candidate 3 1 Administration Sheet 1 Ensure that the computer is ready for the candidate to sit down and begin their assessment session see administration section of user manual 1 April 2004 Page 9 Selby amp Mills Limited Adaptive Ability Tests 2 Ensure that there is some rough paper and two pens or pencils A calculator should be provided if the candidate is to complete the Numeric Ability Test 3 Settle the candidate in front of the screen and make sure they can read the screen perfectly clearly If the candidate needs glasses they may wish to fetch them at this point 4 Once the candidate has been settled say These test s are test s of ability During this session you will complete the following tests THEN NAME THE TESTS THAT HAVE BEEN SELECTED Full instructions are given at the beginning of each test on the computer screen You move from one screen to the next by pressing the return key DEMONSTRATE WHERE The instructions will indicate how to answer and they will also give examples of
22. ble until some time in the future Such criteria may be ratings of an individual for example from a supervisor or they may be a performance measure for example sales figures Concurrent validity is a cruder method of assessing a criterion related validity In this instance the test is administered to a group who are already 1 April 2004 Page 33 Selby amp Mills Limited Adaptive Ability Tests in post Typically the sample would be split between good and poor performers The validation would assess how well the test scores fit the job performance of these individuals The problem with concurrent validity studies is that the sample which is assessed will not be typical of the sample upon which the tests would normally be used A sample of individuals who are already working in a job will already have undergone some self selection e g individuals who were not happy in the job would have left and all the sample will be higher performers by virtue of having been in the job for a period of time D Construction Validity This type of validity is aimed at relating test scores to the underlying characteristic or trait that is being measured This will generally be done by relating the test to a measure of behaviour where the construct is thought to be an important variable For example a test that measures a certain skill would be correlated with a behaviour which could be regarded as typical for that skill Here one needs to consider
23. candidate s interests My mother tongue isn t English The general level of language in the tests is at the 11 12 year level of difficulty The language ability tests is aimed at assessing the range and clarity of an individual s vocabulary This is because this skill is an important requirement for success in the role for which you are being considered If you genuinely do not wish to complete the test you must realise that we will not have as much information on your application as we will have for the other candidates who have completed this test Your application may suffer as a result of this Please complete the test and discuss any difficulties with me afterwards 1 April 2004 Page 12 Selby amp Mills Limited Adaptive Ability Tests 4 SCORING THE ADAPTIVE ABILITY TESTS Each module of the Adaptive Ability Test is automatically scored by the computer The score report that is produced for an individual shows two scores The first score is the raw score which is a direct result of the number of questions which the individual answered correctly The second score on the score report is a STEN score which is an interpretation of that individual s raw score against a selected norm group See below The STEN score is also represented diagramatically as a bar chart on the score report For the numeric and language modules the score report consists of a single pair of scores the raw score and the STEN score For the administrative
24. cells have produced some misleading results and therefore in the case of block 1 level 3 in the numeric and language module estimates have been made to maintain consistency across levels The level 0 items were taken by a different sample and were not comparable with the occupational data These estimated figures will be replaced as further data becomes available The average measuremeters are used to adjust the scores prior to comparison with appropriate normative groups In each case a sub score for each block at 1 April 2004 Page 20 Selby amp Mills Limited Adaptive Ability Tests each level is multiplied by the measuremeter for that block The results of each block are then added to produce a final score TABLE 6 LOCATING LEVEL MEASUREMETERS Language Measuremeters Level 0 Level 1 Level 2 Level3 1 0 22 3 4 3 0 1 0 2 1 3 1 4 3 2 0 3 0 3 2 4 6 Total 4 0 7 3 9 7 11 9 Average 1 0 24 3 2 3 9 Locate items 0 65 Measurements 3 0 Total Average Locate items 0 62 Measurements 3 0 Estimated figures Administrative Number Measuremeters Level 1 Level 2 Level 3 0 5 2 8 1 0 3 7 1 5 Address Measuremeters Level 1 Level 2 Level 3 0 5 1 8 1 0 2 2 1 5 Codes Measurements Level 1 Level 2 Level 3 0 5 3 4 1 0 3 7 1 5 ndicates estimate figures Figures in brackets are actual Measuremeters levels are adj
25. e Adaptive Ability Tests TABLE3 TEST RETEST CORRELATIONS OF THE ADAPTIVE ABILITY TESTS 5 COMPOSITE GROUP Second Administration Numeric Language Administrative First Numeric Administration Language Administration 1 April 2004 Page 17 Selby amp Mills Limited Adaptive Ability Tests 8 8 1 8 2 BACKGROUND AND DEVELOPMENT General The Adaptive Ability Test modules have been developed using the results of more than 86 000 tests over the past ten years in the regular and territorial armies The Regular Army has for many years used aptitude tests to select and allocate personnel to the many specialisations available to entrants The Army Entrance Tests used at present have been developed over a number of years and many thousands of recruits have provided validation information A parallel version of these tests were designed to enable the TA to recruit at the same level as the Regular Army Items included in the Adaptive Ability Tests were developed as part of the above project and trialled with 18 000 applicants prior to inclusion in the Adaptive Ability Tests Further studies were carried out in schools and UK companies to locate the levels of difficulty and examine their use in industry and commerce The Army tests were presented in pencil and paper format Further studies were set up to standardise the tests to be administered and scored by computer Items and the general format of the tests have been
26. er will present them with a set of questions at a higher level of ability If the candidate is answering many questions wrongly or is taking a long time the computer will present them with questions at a lower level of ability As the test session proceeds the performance level is re assessed by the computer on a number of occasions If the performance of the candidate is consistent then the tests will automatically time out If it appears that the candidate is operating at a particularly high level of ability or at a particularly low level of ability the computer will present more items This ensures that the computer will present more items This ensures that the tests discriminate well at either end of the ability spectrum i e the tests discriminate well at the tails of the normal curve 1st April 2004 Page 3 Selby amp Mills Limited Adaptive Ability Tests B There is no need for Alternative Forms Paper and pencil ability tests often require that different forms of a test are used depending on what level of ability the sample has that is being assessed The Adaptive Ability Tests ensure that the individual administering the test does not need to do their own subjective analysis of their candidates ability level before the test has even begun Consequently they do not run the risk of ceiling or floor effects occurring because they have used the wrong form The Adaptive Ability Tests ensure that the whole range of ability is assessed
27. ests The relationship between the Adaptive Ability Tests and other tests are given in the tables below Initial data is given for three of the Personnel Test Battery tests produced by Saville and Holdsworth Ltd in the UK They are A PTB Numerical Computation NP2 This test measures the four basic arithmetic operations of addition subtraction division and multiplication Items include fractions decimals and percentages but avoids more complex mathematical areas such as geometry and algebra B PTB Verbal Meaning VP5 The VP5 is primarily concerned with the meaning of words and the relationship between them The vocabulary is non specialist and is aimed in a broad sense at verbal communication skills it sum eie MMMM s MMMM H n ll 1 April 2004 Page 35 Selby amp Mills Limited Adaptive Ability Tests PTB Checking CP3 This test was designed to measure speed and accuracy in checking written information The task requires the candidate to check a printed list again a hand written list and to spot any errors between the two lists The list contains words numbers letters and symbols TABLE 15 Correlations with other tests Group General Populations Other Test Correlation Coefficient Numeric General Population Administrativ e 11 3 11 3 1 NB the decimal point has been omitted Relationship between the Adaptive Ability Tests and Educational Examin
28. esults of all the even numbered questions may be compared with the results of all the odd numbered questions Different splittings of the test may result in slightly different estimates of consistency of an individuals performance on each item of the test are often used 9 1 Standard Error of Measurement When interpreting an individual s test score knowing that the test has reliability of 0 8 does not help in deciding for example if a borderline score should fall in one direction or another The Standard Error of measurement is a statistic which helps to make this decision The Standard Error of measurement provides a helpful way of interpreting scores Each score should be thought of as having a margin of error above and below which takes account of the possible error of the scores This band of uncertainty can be quantified by testing a sample of individuals who are like those you wish to test and estimating the extent to which their scores varied 1 April 2004 Page 28 Selby amp Mills Limited Adaptive Ability Tests The Standard Error of Measurement SEm is estimated by using the formula SEm SD 1 where SD is the Standard Deviation and is the Reliability DIAGRAM 6 2SEm 1SEm 6 Range of possible scores 0 5 10 In the example above one could imagine that an individual has taken a test and achieved a score of 6 From the standard deviation and reliability of this test as calculated through tests on numerous other
29. he items chosen to represent the quality being measured To do this one needs to use parallel forms of a test In the parallel forms procedure two tests that are equivalent are administered to the same candidates The tests should be equivalent in the sense that they contain similar items of equal difficulty but not the same candidates Unlike the test re test procedure the parallel forms procedure takes into account error variance produced by using different sample of items The most desirable way of calculating reliability would be to correlate the scores obtained from a test form on one particular administration with a parallel form administered to the same group on a subsequent occasion In this way it is possible to assess errors due to the sample of items used and those due to the different conditions of administrations B Internal Reliability Internal reliability assessed error that is produced due to the specific set of item that the test contains Internal reliability can be assessed using parallel forms It should be noted however that parallel forms are expensive and frequently difficult to construct Typically internal consistency is assessed by splitting the test into two halves and comparing results on each of the split halves by correlation Here one can consider each half as being a parallel form The test results can be split in various ways For example the first half may be correlated with the second half or the r
30. it half method designed by Nuttall and Skurnik The method compares favourably with the Kuder Richardson Formula 20 KR20 method and provides additional statistics on each item to enable the comparisons of scores between the 3 Levels of the test The results of the item analysis of the pencil and paper version are given in Table 4 and the Level 0 schools analysis in Table 5 below TABLE 4 PENCIL AND PAPER ITEM ANALYSIS RESULTS LEVELS 1 2 amp 3 Sample Items Sample Mean Standard Measure Reliability Standard Error Sent Size Deviation ments Coefficien of Measurement Split t Languagel Entry Languagel Managers Languagel All Language2 Managers Language3 Graduates Numeric Numeric Numeric Numeric Numeric Number Number Number Number Number Address Address Address Address Address Codes Codes Codes Codes Codes Entry Managers All Managers Graduates Administrative 1 All 1 Entry 1 Managers 2 Managers 3 Managers 1 1 Entry 1 Managers 2 Managers 3 Managers 1 1 Entry 1 Managers 2 Managers 3 Managers C gt gt ooo corrected using the Spearman Brown formula st April 2004 Page 19 Selby amp Mills Limited Adaptive Ability Tests TABLE 5 SCHOOL ITEMS ANALYSIS RESULTS LEVEL 0 Standard Measure Reliability Standard Error Deviation ments Coefficient of Measurement Language Language Bicester Language St Mary s N
31. l Each block consisted of a subtest of 15 items at each of the three levels of supervisory tasks and despite the suitability of the Nuttall technique for small samples results below a sample size of 10 must be considered to be unreliable 1 April 2004 Page 23 Selby amp Mills Limited Adaptive Ability Tests TABLE7 ADAPTIVE ABILITY TESTS THE INTERNAL CONSISTENCY OF TEST SCORES COMPOSITE OPERATIVE CLERICAL SUPERVISING AND MANAGEMENT STAFF N 260 LANGUAGE MODULE ITEM ANALYSIS LEVEL BLOCK SAMPLE MEAN STANDARD RELIABILITY STANDARD ERROR SIZE DEVIATION COEFFICIENT OF MEASUREMENT NUMERIC MODULE ITEM ANALYSIS LEVEL BLOCK SAMPLE MEAN STANDARD RELIABILITY STANDARD ERROR SIZE DEVIATION COEFFICIENT OF MEASUREMENT 7 9 1 68 51 2 51 85 3 8 2 18 18 7 8 1 63 4 0 2 33 64 2 4 1 54 Sample too small 19 3 5 171 19 3 2 2 29 Time trials were carried out to establish cut off times for the locator items of the test The results of these trials are shown in the Table 8 1 April 2004 Page 24 Selby amp Mills Limited Adaptive Ability Tests TABLE 8 LOCATING LEVEL TIME STUDY TIMES IN SECONDS LANGUAGE LEVEL SAMPLE SIZE STANDARD DEVIATION 47 65 40 45 34 74 NUMERIC LEVEL SAMPLE SIZE STANDARD DEVIATION 107 56 159 87 152 40 ADMINISTRATIVE TEST SAMPLE SIZE STANDARD DEVIATION NUMBERS 50 76 ADDRESSES 74 66 CODES 58 94 8 5 Standardisation of the Tests A normative
32. les are highly predicative of work performance on paper work and other administrative tasks The three sections in the Administrative Ability Module are NUMBERS ADDRESSES CODES 1 April 2004 Page7 Selby amp Mills Limited Adaptive Ability Tests Again they are structured to evaluate across a range of difficulty This includes the lower half of the ability range which might be associated with manual workers or operatives Through ability levels which would be typical of clerical and supervisory staff Up to the top end of the ability range and scores within this level represent considerable performance skill The three sections within the administrative ability module provide the user with a breakdown of that candidate s performance on administrative tasks This will give an indication of where specific aspects of an individuals skill may lie For example assessment for an accountancy position will require a high degree of numeric checking whereas other aspects may be of less importance It should be noted that these sections are not discrete An individual will only receive a score on these scales as subsets of the administrative test module These sections should not be given equal weightings as the three main modules they are three subsections of the administrative module 1 April 2004 Page 8 Selby amp Mills Limited Adaptive Ability Tests 3 ADMINISTRATION The computerised version of the Adaptive Ability Tests
33. ll be assessing their numeric ability Face validity is important with respect to how the candidate deals with the tests and the thoughts they have during the test sessions If a candidate perceives a test as having low relevance this can lead to lower motivation and also scepticism B Content Validity This is a measure of whether the test items are representative of the domain that is being measured For example a language ability test containing only items on spelling would be poor content validity if the purpose of the test was to measure general language ability If the contents of the test are well designed then the candidates responses to the items on the test would be representative of what their responses would be to the domain of situations sampled by the test C Predictive and Concurrent Validity Both these types of validity are concerned with the relationship between the test scores and a criterion measure a standard or variable against which test performance can be evaluated Typically in occupational testing the criterion will often be successful in the job Needless to say this approach requires us to be able to define what can be regarded as success in the job and also that we are able to measure this accurately Predictive validity takes a longitudinal approach to this and is the most relevant to the selection process In this case test scores are correlated with some measure of the criterion which will not become availa
34. nsistency The coefficient is calculated by comparing two sets of scores which are achieved by an individual The source of these scores will vary depending upon which source of inconsistency is being assessed Test Retest Reliability This is a measure of the consistency of the scores that a group of individuals achieve if they complete the tests of two different occasions This particular type of reliability is aimed at assessing the influence of external factors which may affect a candidates performance This includes factors such as the conditions under which the test is taken and how an individual feels on the particular day they take the test The resulting co efficient therefore gives information the stability of test scores over time It should be noted that the independent variable in test retest reliability is the length of time that lapses between each test administration The results may therefore be contaminated by factors such as memory and motivation among the group However in the Adaptive Ability Tests because they are adaptive tests the questions the candidate answers will change if they are performing at a higher or lower level of ability This ensures that memory will at best only have a marginal effect on a particular candidates performance 1 April 2004 Page 27 Selby amp Mills Limited Adaptive Ability Tests This particular form of reliability does not assess spurious variance which may occur due to t
35. s with speed and relative accuracy up to 1 April 2004 Page 6 Selby amp Mills Limited Adaptive Ability Tests about O or GCSE Level C High scores on this type of test suggest that a person will cope with numeracy generally used by managerial and supervisor tasks in non technical settings It is a key managerial skill and a key deficit among managers and graduate candidates for employment This ensures that the module will assess across the whole range of ability although not the specialist maths skills of a graduate mathematician and that the module is appropriate to all levels of occupational use Diagram 1 below indicates the range of individuals that can be assessed using the language and numeric modules DIAGRAM1 THE RANGE OF INDIVIDUALS WHO CAN BE APPROPRIATELY ASSESSED BY THE LANGUAGE AND NUMERIC MODULES OF THE ADAPTIVE ABILITY TESTS Junior Secondary College School School Students Undergraduates Graduates Clerical Supervisory X Line Senior Staff Staff Managers Managers Ability Numeric Level Language 24 Administrative Ability This module is split up into 3 sections The administrative ability sections are performance tests which measure the speed and accuracy with which a person approaches clerical tasks Though this type of measure is predominantly used to select clerical and administrative personnel the ability to organise and perform well on such tasks is a requirement at all levels in modern companies The modu
36. score less well when compared to a group of graduates For this reason one should always ensure that the norms being used are relevant to a given situation Of the various norm systems that are available the Adaptive Ability Tests uses one of the standard score systems know as the Standard Ten STEN The standard score gives an indication of how far above or below the mean a score is This interpretation is done in relation to the normal distribution curve See Diagram 3 This standardisation of the scores means that a candidate s performance can also be compared between the different ability tests in terms of norms That is to say that an individual that scores a STEN of 7 against a graduate norm on numeracy and a STEN of 4 against a graduate norm on language can be compared directly across these two modules i e that individual is scoring well above average compared with a graduate group on numeracy and below average compared with a graduate group on language Another advantage of using the normal distribution as a basis for test norms is that the standard deviation has a precise relationship to the area under the curve This in turn means we are able to say what percentage of the norm group would achieve a given STEN SCORE Diagram 3 shows the cumulative percentage of scores So for example if an individual scores a STEN of 4 we would expect 30 of the norm group to score at or below this level 1 April 2004 Page 14 Selby amp
37. study was carried out and cut off scores were calculated based on the performance of the three groups in the sample managerial supervisory and operative These cut off scores are shown in Table 9 for the paper and pencil versions of the test TABLE9 NORMATIVE DATA CUT OFF SCORES MANAGERIAL SUPERVISOR OPERATIVES LANGUAGE NUMERIC ADMINISTRATIVE NUMBERS ADDRESSES CODES Scores marked with a indicate that this is not a management cut off score rather it reflects a very high level of performance Individuals who score high on the other tests but low on the administrative module are usually preoccupied with precision In the development of the Differential Aptitude ILLI LL A 1 April 2004 Page 25 Selby amp Mills Limited Adaptive Ability Tests Test Psychological Corp a group of low performance high aptitude scorers were retested with new instructions They were asked to work as fast as possible without worrying about precision members of the group improved their scores radically without any appreciable loss of accuracy You would expect that a low score in these cases has more to do with the work style of the individual rather than the level of performance Diagram 5 shows graphically the cut off scores giving an indication of the type of profile you might expect in each case from each of the three groupings Precise normative information is included in the
38. t using measuremeters a more accurate method of determining whether a score on one test is more significant than a score on another test would be to combine the SEm on both The Standard Error of Measurement indicated to what extent a test score is a true score for that individual Similarly the standard error of difference indicates the extent to which given two scores on two different test this score difference is an indication of true ability differences Thus for example if an individual gets a score of 6 on the first test and a score of 8 on the second test and the SEd is 2 score units then you can be sure that 68 of the time there is a true ability difference So for example we could use the following formula to give the SEd in score units SEd SEm 12 SEm 22 The resulting SEd gives us a score value to show the precise difference between a score obtained on one test and the score obtained on one test and the score on another As with the SEm 1 SEd suggests that there is a 6876 probability that a score obtained on one test is different from a score obtained on another test 2 SEds provide a probability of 9576 that a true ability difference exists The Standard Error of Differences on the Adaptive Ability Tests are given in Table 10 The lower the score the stronger the relationship between the tests TABLE 10 STANDARD ERROR OF DIFFERENCE BETWEEN TESTS LANGUAGEI NUMERICI LANGUAGE2 NUMERIC2 NUMBERS1 ADDRESS1 CODES1 NUMBE
39. test module the score report shows the level of attainment in each of the three sections of that module as well as an overall score on the administrative ability module The score report therefore looks as follows DIAGRAM 2 SAMPLE SCORE CHART Score Chart Report Report Date 21 03 2004 Administrative Ability Score STEN 1 2 3 4 5 6 7 8 9 10 OVERALL 152 8 m Numbers 30 5 55 5 28 5 EM Codes 94 10 Raw Score STEN 1 2 3 4 5 6 7 8 9 10 Language Ability 46 5 J M Raw Score STEN 1 2 3 4 5 6 7 8 9 10 Numeric Ability 61 7 m Norm group used General Population This is THE END of the report 1 April 2004 Page 13 Selby amp Mills Limited Adaptive Ability Tests 5 INTERPRETATION THE USE OF NORMS Normative data gives us guidelines against which an individual s scores can be interpreted To know that an individual s raw score is thirty seven does not tell us whether this can be regarded as good bad or indifferent To interpret a given score correctly we need to compare an individual s scores against the scores of a large group of people Such a group would be called a norm group Norm groups can be drawn from various populations For example graduates or clerical staff Obviously one can reach entirely different conclusion depending with which norm group an individual is compared One would expect someone who scores well on the numeric ability module compared with a group of clerical staff to
40. the types of questions you will have to answer Do any rough work on the paper provided and use the calculator if you wish Each test is automatically timed The time will vary but generally each test will last between 10 and 20 minutes Do not worry if the test finishes while you are in the middle of a question Timing does not start until you are presented with the first question of the test Please read the instructions carefully When you have completed the tests please inform the administrator Are there any questions before you begin ANSWER ANY QUESTIONS THAT THE CANDIDATE HAS SEE SECTION 3 2 AND THEN SAY You may begin now 3 2 Questions and Answers for Administrators Do I lose marks for incorrect answers No you do not lose marks if you answer any of the questions incorrectly How well you do on each test will depend on a combination of three factors Firstly the number of questions that you answer correctly Secondly the difficulty level of the questions that you answer for example a hard question will be worth more marks than an easy one The final factor is the length of time it takes you to answer the questions that you attempt The computer will not add or subtract anything from your score if you answer a question incorrectly How long will the tests take DILE D 8 1 April 2004 Page 10 Selby amp Mills Limited Adaptive Ability Tests Each of the tests takes between 10 and 20 min
41. umeric Numeric Bicester Numeric St Mary s 8 3 Measuremeters The levels in the Ability Tests have been derived by using the measuremeter statistic described by Nuttall and Skurnik 1974 The statistic provides an indication of how reliable the test is The higher the measuremeter the higher the reliability of the test In addition a further characteristic of this statistic is the relation of one measuremeter to another test taken by a group producing four measuremeters could be described as twice as reliable as a test taken by that same group with two measuremeters This linearity of scale is not possessed by the reliability coefficient and provides a pragmatic method of linking the three levels of the Adaptive Ability Tests Nuttall and Skurnik s method of converting scores to grades by dividing by the appropriate measuremeter was adapted to produce a method of weighting scores on more difficult tests to produce higher scores This method has been adopted as an interim while development of a fully tailored testing system is developed using Rasch item scaling or other appropriate probabilistic models Table 6 shows the measuremeters for the three modules of the Adaptive Ability Tests Please note that the table gives results for each of the blocks that were developed as part of the computerised version see below The table shows the measuremeter calculations for establishing levels of difficulty The low sample sizes in some
42. umeric Ability tests For the Administrative Ability tests there was a problem using these blocks of items When the candidate changed blocks they also had to receive additional instructions As a result they would actually achieve a lower score by being unable to maintain performance at the new level Therefore for the Administrative Ability tests once the candidate is allocated to a particular level they receive all the items at that level For the Numeric Ability and Language Ability tests this partial tailoring of testing was trialled in the next stage Diagram 4 illustrates the structure of the three test modules 1st April 2004 Page 22 Selby amp Mills Limited Adaptive Ability Tests DIAGRAM 4 STRUCTURE OF THE MODULES LANGUAGE ABILITY STRUCTURE Level 3 Level 2 Level 1 5 Level 0 Locate questions Block 1 Block 2 Block 3 Level 3 Level 2 Level 1 5 Level 0 Locate questions NUMBERS SECTION Level 3 Level 2 Level1 4 Locate questions ADDRESSES SECTION Level 3 Level 2 Level1 4 Locate questions Test questions CODES SECTION Level 3 14 WORDS Level 2 4 WORDS 14 WORDS Level 1 25LETTERS 14 WORDS Locate questions Test questions A sample of 260 operatives clerical supervisory and management staff were tested on the computerised version The results of the item analysis for language and numeracy are given in Table 7 The tests were analysed by block of items at each leve
43. usted _ _ 1 April 2004 Page21 Selby amp Mills Limited Adaptive Ability Tests 8 4 Computerised Version From these studies it was possible to construct a computerised version of the tests where each could be sub divided into three levels in the case of the Administrative Ability tests and four levels in the case of Language and Numeric Ability tests Within these levels blocks of items were separately tested in groups of 15 to establish whether individuals could be re allocated to levels during testing Individuals are allocated to a given difficulty level on the basis of their performance on a series of locator questions at the beginning of the test The sample used to locate level 0 was different to that used to locate levels 1 2 and 3 This means that currently the locator questions operate at level 1 2 and 3 The starting algorithm will be changed as data becomes available to include locator items at level 0 If a person was allocated to Level 1 at the beginning of the test and then performed better than the criterion for that level then that person could be moved up a level for the second block of items Similarly if it was then clear that he was struggling at the new level he would drop down for the final block of items In practice this system was only practical for the Language Ability and N
44. utes to complete The timing is carried out automatically by the computer This means that any test may finish when you are halfway through a question try not to let this bother you Also you may find that different people will take different amounts of time to complete the test This is because sometimes the computer will take longer to identify the candidates level of ability This is especially true for candidates who answer erratically How many questions will I have to do This will depend on how quickly you answer the individual questions Generally you will have to answer between 30 and 40 questions If you are answering quite cautiously and therefore slowly you may complete less questions in the time limit allowed However you should remember that the questions vary in difficulty This means that if you answer fewer of the more difficult questions you could get the same score as someone answering a greater number of the less difficult questions Speed and accuracy are both taken into account Are the tests valid for me These tests have been developed as the result of extensive trials involving thousands of people These have included secondary school children university students graduates clerical staff senior managers and others The tests have been specifically designed to measure the abilities of all types of people of working age The reason that you have been asked to complete these tests is that they have been found to be an
45. wered with a paper and pencil test 2 2 Language Ability The language module was designed to test the range and clarity of a person s vocabulary as well as their ability to use language Analogue tests of this type are highly correlated with verbal reasoning intelligence tests Such tests are used to predict successful performance in examinations and training courses which have a high verbal component High scores on this test would be essential for positions which require sophisticated communication skills Similarly minimum scores are required for operative or supervisory roles where precision is required in giving and receiving instructions It is to be remembered however that a low score on this test is not an indicator of low intelligence Many scientists and technologists have poorly developed verbal skills whilst possessing high levels of abstract and spatial ability The module covers a range of ability from final year junior or first year secondary school through secondary school up to graduate level 2 3 Numeric Ability The numeric ability module was designed to measure the aspects of numeracy generally required in work situations At lowest difficulty level simply four rule numeracy is covered In the higher difficulty levels the emphasis is on the individual s ability and confidence when using numbers or formulae Though little numeric knowledge is required the module examine a person s ability to estimate and solve problem
Download Pdf Manuals
Related Search