Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
27 Cards in this Set
- Front
- Back
|
Test Construction
Glossary Alternate Forms of Reliabity |
Mthd 4 estmtng a tst's rliablty tht ntails admnstrng 2 forms of th tsts 2 th sam grp of xamnees nd crrltng th 2 sets of scors. Forms cn b admnstrd at abt the sam tme (coeffcnt of equvlnc) or at dffrnt tms (coeffcnt of equvlnc nd stblty). Cnsdrd by som xprts 2 b th bst (most thoro) mthd 4 assng reliablty.
|
|
|
Test Construction
Glossary Classical Test Theory |
Thry of msmnt tht rgrds obsrvd variblty n tst scors as rflctng two cmponents: tru dffrnc btwn xamnees on th attrbut(s) measrd by th tsts nd th effcts of th msmnt (random) error. Reliablty is a meas of tru scor variblty.
|
|
|
Test Construction
Glossary Coefficient Alpha/KR-20 |
Mthd 4 assng ntrnl cnsistncy rliablty tht provids an ndx of avrg intr-itm cnsistncy. Kuder-Richardson Formula 20 (KR-20) cn b usd as a substitue 4 coeffcnt alpha whn tst itms r scored dichotomusly
|
|
|
Test Construction
Glossary Constrct Validity/Convergent & Discrimant |
Tn xtnt 2 wch a tst meas th hypothtcl trait (constrct) it is ntnded 2 meas. Mthds 4 estblshng cnstrct valdty nclud crrltng tst scors w/scors on meas tht do & do nt meas th sam trait (convergent & dscriminant validty); condctng a factr analysis to asess th tst's fctrl valdty; dtrmng if chngs n tst scors rflct xpctd dvlpmntl chngs; nd seeng if xprmntl manipultns hv th xpctd mpct on tst scors.
|
|
|
Test Construction
Glossary Content Validty |
th xtnt 2 wch a tst adqutly sample th doman of nfo, knwldg, or skl tht it purprts 2 meas. dtrmn prmrly by "xprt judgmnt." Most mprtnt 4 achevmnt nd job sampl tst.
|
|
|
Test Construction
Glossary Criterion Contamination |
Rfrs 2 bias ntrducd n2 a prsn's critrion scor as a rslt of th knwldg of th scorer abt scorer's prfrmnc on th prdctr. Tnds 2 artificly nflat th r/s btwn prdctr nd criterion.
|
|
|
Test Construction
Glossary Criterion-Referencd Interpretation |
Ntrprtatn of a tst scor n trms of a prespcfd std; i.e., n trms of % of cntnt crrct (%scor) or of predctd prfrmnc on an xtrnl criterion (e.g., regrssn equtn, xpctncy tabl).
|
|
|
Test Construction
Glossary Criterion-Related Valdity / Concurrent & Predictive |
Th typ of valdity tht nvolves dtrmng th r/s (crrltn) btwn th predctr nd th criterion. Th crrltn coeffcnt is rffrd 2 as th criterion rltd valdity coeffcnt. Criterion-rltd valdity cn be either cncurrent (predctr nd criterion scors obtained at about the sam time) or predctiv (predctr scors obtnd b4 criterion scors).
|
|
|
Test Construction
Glossary Cross-Validation & Shrinkage |
Procss of re-assng a tst's criterion-rltd validty on a nu smpl 2 chk th gnrlzblty of th orgnl valdty coeffcnt. Ordnrly, th valdty coeffcnt "shrnks" (bcoms smaller: on crss-valdtn b/c th chnc fctrs opertng nn th orinal sample r nt all prsnt n th crss-valdtn sampl,
|
|
|
Test Construction
Glossary Factor Analysis |
A multivariate statstcl tchq usd 2 dtrmn hw mny fctrs (cnstructs) r needed 2 accnt 4 th ntrcrrltns amng a set of tsts, subtsts, or tst itms. Fctr analysis cn b usd to assess a tst's cnstruct validty by ndctng th xtnt 2 wch th tsts crrlts w/fctrs tht wud nd wud nt b xpctd 2 crrlt with. From th prspctv of fctr analys, tru scor variability consists of communality nd spcfcity. Fctrs idntfd n a fctr analys cn b eithr orthgonical or oblique.
|
|
|
Test Construction
Glossary Factor Loadings & Communality |
N factr analys, a fctr lodng is th crrltn btwn th tst (or othr varibl ncluded n th analys) nd a fctr. Th communality 4 th factr analys (i.e., by th identfied fctrs).
|
|
|
Test Construction
Glossary Incremental Validity/True Positives, False Positives, True Negatives, False Negatives |
th xtnt 2 wch a predctr ncrese dcsn-mkng accurcy. Calcultd by sbtrcrng th bse rat from th pstv hit rat. trms 2 hv lnkd w/ncrmentl valdity r predctor nd criterion cutoff scores; tru nd fals positvs nd tru nd fals ngtvs. Tru pstvs r thos who scord hi on th prdctr nd th criterion; fals postvs scord hi on th prdctr bt lo on th crterin; tru negtvs scord lo on th predctr nd th criterion; and fals ngtvs scored lo on th prdctr bt hi on th criterion.
|
|
|
Test Construction
Glossary Item Characteristc Curve |
Whn usng itm rspns thry, an itm chrctrstc curv (ICC) is cnstrctd 4 ea itm by plottng th proportn of xamnees n th tryout sampl hu answerd th itm crrctly agnst eithr th totl tst scor, prfrmnc on an xtrnl criterion, or a mthmtcly-derived est of a latent ability or trait. Th curv prvids nfo on th r/s btwn an xamnees's lvl on th ablty or trait measrd by th tst nd th probablty tht xamnee wil rspond to th itm corrctly.
|
|
|
Test Construction
Glossary Item Difficulty |
An itm's dffclty is clcultd by dvdng th # of ndvdls hu answrd th itm crrctly by th totl # of ndvdls; rngs n valu frm 0 (vry dffclt itm) to 1.0 (vry ez itm). N gnrl, an itm dffclty ndx of .50 is prfrd b/c it mxmzs dffrntatn btwn ndvdls w/hi nd lo ablty nd hlps nsur a hi rliblty coeffcnt.
|
|
|
Test Construction
Glossary Item Discrimination |
Itm dscrmntn rfrs 2 th xtnxt 2 wch a tst itm dscrmnts (dffrntiats) btwn xamneex hu obtn hi v lo scors on th ntir tst or on xtrnl critrion. Th itm dscrmntn ndx (D) rngs frm -1.0 to +1.0. If all xamnees n th uppr grp nd non n th lwr grp answrd th itm crrctly, D is +1.0; if non of th xamnees n th uppr grp na all of th xamnees n th lwr grp answrd th itm corrctly, D= -1.0.
|
|
|
Test Construction
Glossary KAPPA Statistic |
A crrltn coeffcnt usd 2 asses ntr-rtr rliablty.
|
|
|
Test Construction
Glossary Multitrait-Multimethod Matrix. |
A systmtc wy 2 orgnz th crrltn coeffcnt obtnd whn assng a measr's cnvrgnt nd dscrmnt vldty (wch n trn, prvids evdnc of cnstrct vldty). Rquirs measrng at least 2 dffrnt traits usng at least 2 dffrnt methds 4 ea trait. Terms 2 hv lnkd w/multitrt-multimthd matrx r monotrait-monomthd, momotrait-heteromthd, heterotrait-monomthd, nd hetertrait-heteromthd coeffcnts.
|
|
|
Test Construction
Glossary Norm-Referenced Interpretation |
Ntrprtn of an xamnee's tst prfrmnc rltv 2 th prfrmnc of xamnees n a nrmtv (stndrdztn) sampl. %tile rnks nd std scors (e.g., z-scors nd T scors) r typs of nrm-rfrncd scors.
|
|
|
Test Construction
Glossary Orthogonical & Oblique Rotation |
N fctr analys, an orthgncl rotatn of th idntfd fctrd prducs uncrrltd fctrs, whl an oblq rotatn produces crrltd fctrs. Rotatn is don 2 smplfy th ntrprtn of th idntfd fctrs.
|
|
|
Test Construction
Glossary Relationship Between Reliabilty & validity |
Rliablty is a ncssry bt nt sffcnt cndtn 4 vldty. N trms of critrion-rltd vldty, thvldty coeffcnt cn b no grtr th th squr root of th product of th rliblties of th predictr nd critrion.
|
|
|
Test Construction
Glossary Relevance |
N tst cnstrctn, rlvnc rfrs 2 th xtnt 2 wch tst itms cntrbut 2 achevng th statd gols of tstng.
|
|
|
Test Construction
Glossary Reliability/Reliability Coefficient |
Rliblty rfrs 2 th cnsistncy of tst scors; i.e., th xtnt 2 wch a tst measrs an attrbut w/o bng affctd by rndm flctuatns (memnt error) tht produces ncnsistncies ovr tme, acrss tme, or dffrnt forms. Mthds 4 estblshng rliblty nclud tst-rtst, altrntv forms, slit-hlf, coeffcnt alpha, nd ntr-rtr. Most produc a rliblty coeffcnt, wch is ntrprtd dirctly as a meass of tru scor varblty - e.g., a rliblth of .80 ndicats tht 80% of varblty n tst scors is tru scor varblty.
|
|
|
Test Construction
Glossary Split-Half Reliability / Spearman-Brown Formula |
Splt-hlf rliblty is a mthd 4 assng ntrnl cnsistncy rliblty nd nvolvs "slttng" th tst n hlf (e.g., odd- vrs evn-#d itms) nd crrltng xamnee's scors on th two hlvs of th tst. Th splt-hlf rliblty coeffcnt is usuly crrctd w/th Spearman-Brown frmula, wch estmts wht th tst's rliblty wud b if it wer bsed on th full lngth of th tst.
|
|
|
Test Construction
Glossary Standard Error of Estimate / Confidence Interval |
An ndx of eror whn prdictng critrion scors from prdictr scors. Usd 2 cnstrct a cnfdnc ntrvl arnd an xamnee's prdictd critrion scor. Its mgntud dpnds on two fcts: th critrion's std dvtn nd th prdctr's vldty coeffcnt.
|
|
|
Test Construction
Glossary Standard Error of Measurement / Confidence Interval |
An ndx of msmnt eror. Usd 2 cnstrct a cnfdnc ntrvl arnd an xamnee's obtnd tst scor. Ts mgntud dpnds on two fctrs: th tst's std dvtn nd rliblty coeffcnt.
|
|
|
Test Construction
Glossary Test Length/Range of Scores |
A tst's rliblty cn b ncresd n svrl wys. One wy is 2 ncres th tst lngth by adng itms of simlr cntnt nd qlty. Anthr is 2 ncres th heterognty of th sampl n trms of th attrbut(s) measrd by th tst, wch wil ncres th rng of scors.
|
|
|
Test Construction
Glossary Test-Retest Reliabilty |
A mthd 4 assng rliblty tht nvolvs admnstrng th sam tst 2 th sam grp of xamnees on two dffrnt occsns nd crrltng th two sets of scors. Ylds a coeffcnt of stbilty
|
