語言項目中的測試：英語語言測評綜合指南（簡體書）

《英語教師職業發展前沿論叢．語言項目中的測試：英語語言測評綜合指南》主要介紹了語言測試中的基本概念，以及如何選用和設計測試類型和題型。介紹測試結果的量化分析和解讀。討論語言測試中的兩大基本要素：測試信度和效度。《英語教師職業發展前沿論叢．語言項目中的測試：英語語言測評綜合指南》的主要框架分為前言、章節內容（1-11章）、答案、術語表、索引、參考文獻六個部分。每章內容的基本結構：先介紹相關理論和概念，進而闡述和討論其在不同情境中的實踐和運用，接著運用Excel工作錶針對本章節的數據分析進行操作指導，然後以問題方式對本章節進行總結和回顧，最後是應用練習。《英語教師職業發展前沿論叢．語言項目中的測試：英語語言測評綜合指南》將理論和實際相結合，緊扣課堂實際和教師需求，針對性強。每章的數據分析都有手把手的實際操作指導，每章後的總結和思考題便於讀者梳理和消化所學知識，設計的練習和所附答案有助於鞏固加強對已學內容的掌握和應用，有助於自學。本書適用面廣，讀者群體可包括教育決策者、教學主管、語言教師、測試命題人員、測試研究人員以及相關學科研究生等。

《英語教師職業發展前沿論叢?語言項目中的測試:英語語言測評綜合指南》將理論和實際相結合，緊扣課堂實際和教師需求，針對性強。每章的數據分析都有手把手的實際操作指導，每章后的總結和思考題便于讀者梳理和消化所學知識，設計的練習和所附答案有助于鞏固加強對已學內容的掌握和應用，有助于自學。《英語教師職業發展前沿論叢?語言項目中的測試:英語語言測評綜合指南》適用面廣，讀者群體可包括教育決策者、教學主管、語言教師、測試命題人員、測試研究人員以及相關學科研究生等。

叢書總序中文導讀PrefaceChapter1TypesandUsesofLanguageTestsTwoFamiliesofLanguageTestsNorm-ReferencedTestsCriterion-ReferencedTestsTypeofInterpretationTypeofMeasurementPurposeoftheTestingDistributionsofScoresTestStructureMatchingTeststoDecisionPurposesProgram-LevelProficiencyDecisionsProgram-LevelPlacementDecisionsClassroom-LevelAchievementDecisionsClassroom-LevelDiagnosticDecisionsWhyaSingleTestCannotFulfillAllFourFunctionsDifferencesinRangesofAbilityDifferencesinVarietyofContentUsingSpreadsheetProgramsinLanguageTestingWhatIsaSpreadsheetProgram?HowWillYouPersonallyBenefitfromUsingaSpreadsheetPrograminThisBook?ReviewQuestionsApplicationExercisesChapter2Adopting,Adapting,andDevelopingLanguageTestsTheoreticalIssuesLanguageTeachingMethodologyIssuesAnExceptionallyShortHistoryofLanguageTestingWhyKnowingaboutTheseMovementsIsImportantTheCompetence/PerformanceIssueTheDiscrete-Point/IntegrativeIssuePracticalIssuesTheFairnessIssueTheCostIssuesEaseofTestConstructionEaseofTestAdministrationEaseofTestScoringInteractionsofTheoreticalIssuesAdopt,Adapt,orDevelop?AdoptingLanguageTestsAdaptingLanguageTestsDevelopingLanguageTestsPuttingSoundTestsinPlaceGettingStartedwithYourSpreadsheetProgramMovingAroundtheSpreadsheetCreatingaSampleSpreadsheetEnteringTestScoreDatatoCreateaSpreadsheetReviewQuestionsApplicationExercisesChapter3DevelopingGoodQualityLanguageTestItemsWhatisaTestItem?GuidelinesforItemFormatAnalysisGeneralGuidelinesReceptiveResponseItemsProductiveResponseItemsPersonalResponseItemsWhyBotherwithItemFormatAnalysis?ReviewQuestionsApplicationExercisesChapter4ItemAnalysisinLanguageTestingNorm-ReferencedItemAnalysisItemFacilityAnalysisItemDiscriminationAnalysisCalculatingItemFacilityandDiscriminationwithYourSpreadsheetNRTDevelopmentandImprovementProjectsCriterion-ReferencedItemAnalysisItemQualityAnalysisCRTDevelopmentandImprovementProjectsRoleofItemFacilityDifferenceIndexTheB-IndexCRTItemSelectionReviewQuestionsApplicationExercisesChapter5DescribingLanguageTestResultsDisplayingDataGraphicDisplayofFrequenciesCreatingGraphsinExcelTMScalesofMeasurement……Chapter6InterpretingLanguageTestScoresChapter7CorrelationinLanguageTestingChapter8LanguageTestReliabilityChapter9LanguageTestDependabilityChapter10LanguageTestValidityChapter11LanguageTestinginRealityAnswerKeysGlossaryIndexReferences

CRITERION-REFERENCED ITEM RNRLYSIS
Recall that a central difference between NRTs and CRTs is that NRTs (typicallydesigned to spread students out in percentile terms for proficiency or placement testingpurposes) are constructed to produce normal distributions, while CRTs (because theyare designed to measure what students know or can do in percentage terms fordiagnostic or achievement purposes) do not necessarily do so. In addition, the itemselection process for developing NRTs is designed to retain items that are well-centered (with IFs of .30 to .70) and spread students out efficiently (the highest IDs areretained). Such items once selected for a revised version of an NRT will generallywork together to provide a normal distribution of scores.
In contrast, CRTs may not necessarily produce scores that are normally distributed.
In fact, a CRT that is designed to measure student achievement might produce scoresthat are predominantly high. Consider an ideal situation in which all the studentsmastered all the objectives in a particular course because they were highly motivatedstudents, the materials were marvelous, and the teacher was, of course, spectacular. Allof those students could reasonably be expected to score 100 percent on the end-of-course criterion-referenced achievement test designed to measure those objectives.Of course, a teacher could create the same effect (that is, everyone scoring100 percent) by writing a final examination that is far too easy for the students.Tocheck for this possibility, the teacher may want to administer the test (or an equivalentform of the test) at the beginning of the course as a diagnostic test. If the studentsperform poorly on the beginning-of-course diagnostic test (pre-test) and score well onthe end-of-course achievement test (post-test), then the teacher can interpret the highscores at the end of the course as legitimate reflections of the students' knowledge orskills, rather than reflections of a test that is too easy for the students.