Development of theSubject Tests

            Development of theSubject Tests" G) \9 v% ^3 Q+ B* G6 ~
Each new edition of a Subject Test is developed by a
% K4 H; n3 O, R  acommittee of examiners composed of professors in the* H- o0 w0 p9 t6 n9 D
subject who are on undergraduate and graduate facul-ties6 O: H2 y/ ^. |
in different types of institutions and in different
9 `, R5 v6 A/ C: nregions of the United States and Canada. In selecting
! D  g& M. v  n8 omembers for each committee, the GRE Program seeks( k! L7 N, R4 {' z. Z0 V% d
the advice of the appropriate professional associations. V6 M: T. I; O
in the subject.3 C0 ^3 L' N+ Q1 _4 a4 }
The content and scope of each test are specified
7 E  Y' T) u$ [) X" _5 Vand reviewed periodically by the committee of exam-iners.
. C& y  ?+ `$ K, Q. ITest questions are written by the committee and
, N! Q; X/ ?# q: }# sby other faculty who are also subject-matter specialists
, F$ b- ]3 v1 cand by subject-matter specialists at ETS. All questions' d) ]: w7 l9 |; T& e" q" G, P) w& P9 t
proposed for the test are reviewed by the committee
) A9 L9 O( w( ^" Vand revised as necessary. The accepted questions are
% r  ^2 H4 {3 }" C) I8 Massembled into a test in accordance with the content
; h0 e$ u  o0 X1 F+ U, Xspecifications developed by the committee to ensure# B9 a, p8 P- m4 {" g
adequate coverage of the various aspects of the field( Q% p% u( L) @0 v0 G' {
and, at the same time, to prevent overemphasis on any. u2 S, S5 r+ p0 h2 j
single topic. The entire test is then reviewed and - r7 Y: o- J1 e* ^
approved by the committee.+ `' ~- O! ~( W4 i
Subject-matter and measurement specialists on the
. t: K% I  F$ f* V1 o9 |ETS staff assist the committee, providing information
. _7 [+ g: W4 H  o/ z' Iand advice about methods of test construction and
: |" p) I: j2 w  }2 lhelping to prepare the questions and assemble the test.3 ]- e2 g+ D  c3 z
In addition, each test question is reviewed to eliminate1 }3 D$ X5 ~5 e7 }
language, symbols, or content considered potentially
3 Y7 `7 g. B0 R* T5 B. Aoffensive, inappropriate for major subgroups of the test-taking
" e/ W7 v0 b8 B; ipopulation, or likely to perpetuate any negative
5 k; E3 s7 X- m+ P0 Qattitude that may be conveyed to these subgroups. The
3 O, ?; x: R5 \5 S' q/ X: Stest as a whole is also reviewed to ensure that the test
0 ]& g; L, K5 d+ V9 w1 G! R8 [0 squestions, where applicable, include an appropriate
$ e& O& B1 K& U2 W/ b/ [) u! s) |balance of people in different groups and different roles.
; O' F, v. @1 ~  \# Y, GBecause of the diversity of undergraduate curricula,
* Z6 O& Y; B$ @% \% nit is not possible for a single test to cover all the$ L+ |% S* ]6 |0 o5 ^" L
material you may have studied. The examiners, there-fore,
/ ?: Z* w$ @* O$ yselect questions that test the basic knowledge and
' @9 a9 ~6 j. g# _4 O- ^! Nskills most important for successful graduate study in
0 u. `  c2 k- p; D- c( H3 c$ u* Ithe particular field. The committee keeps the test& g' Z- V# n0 T( [/ G* D, G( o
up-to-date by regularly developing new editions and
: u! s) O: i0 n2 C5 @revising existing editions. In this way, the test content0 C/ Z  Z0 m. b; `
changes steadily but gradually, much like most cur-ricula.7 p. G, J$ {# V" q6 R
In addition, curriculum surveys are conducted+ f" e3 Y* J6 M* ~0 J  c7 t/ K
periodically to ensure that the content of a test
) j" F; q" G' j/ freflects what is currently being taught in the under-graduate
/ J) b8 n. p5 J4 r, ^curriculum.
* j9 n5 N- U( M+ p+ b/ t8 HAfter a new edition of a Subject Test is first admin-istered,
$ Z2 Z* c% m7 N& [3 G1 B/ bexaminees’ responses to each test question are- A% H* M) K* G; s" j: x7 f2 A: E
analyzed in a variety of ways to determine whether6 d) V  ?5 P9 T5 K2 p, E
each question functioned as expected. These analyses+ G6 ]) j3 W( z: l/ K! Z
may reveal that a question is ambiguous, requires! b+ o+ N8 e" P1 h% ]. l
knowledge beyond the scope of the test, or is inappro-priate% B8 l+ i( A: N) ]
for the total group or a particular subgroup of
! ^2 [4 {: C. L3 Y  b0 i: E* h7 Rexaminees taking the test. Answers to such questions
9 v: d- f2 L0 r: J- Zare not used in computing scores.
) l+ Z0 N& b6 EFollowing this analysis, the new test edition is
# z% B3 @# X9 A1 Vequated to an existing test edition. In the equating
7 F: c: G! o* Vprocess, statistical methods are used to assess the
9 @8 H  L. o( w" u. N# }difficulty of the new test. Then scores are adjusted so
# A$ Z+ q& s. ?% @* Vthat examinees who took a difficult edition of the test
# y9 X- s; ?+ Y4 @" Oare not penalized, and examinees who took an easier& t! O/ G; f( v2 n9 Q% [
edition of the test do not have an advantage. Varia-tions
* a; i* ~) y" C3 W# a8 `9 d+ qin the number of questions in the different, x- C6 {, d4 d% _; ]
editions of the test are also taken into account in
$ ~2 y6 h. [$ U2 g) i6 Kthis process.
8 X; Z$ l- k6 G. RScores on the Subject Tests are reported as three-digit
$ x; [/ u! V1 v* o7 }6 f; d% nscaled scores with the third digit always zero.
9 |& R5 o$ q7 T. XThe maximum possible range for all Subject Test total
8 Y7 j# B% v3 d' k* |scores is from 200 to 990. The actual range of scores for& O4 h) r& I: _* M# i2 @
a particular Subject Test, however, may be smaller. The
  m0 `4 b0 k* p0 f* s0 ?) Y& |maximum possible range of Subject Test subscores is- _; O2 p$ W- J- s9 n1 _+ r* `
20 to 99; however, the actual range of subscores for+ g' [, k5 ~1 g+ q" e. |
any test or test edition may be smaller. Subject Test
5 t* {' B0 I# u  c; Fscore interpretive information is provided in Interpret-ing
& k" c0 t! u( Q1 F; A0 O9 ^Your GRE Scores, which you will receive with
5 O+ ^: k' @; n& Q' ?- ryour GRE score report, and on the GRE Web site at# ?/ e- C' n! C- n4 P- _" Q; n
www.gre.org/codelst.html.
- c# o1 m% u' d. U6 d