简介:口语考试作为主观性考试,其评分准确性在很大程度上会影响考试的信度和整体效度。本研究结合对评分结果的多层面Rasch模型统计分析和对评分员评分依据的定性描述,探讨PETS三级口语考试中常见的评分误差类型、程度和造成误差的可能原因,并据此提出了改进评分标准,评分流程和评分员培训的建议。
简介:AIM:Todemonstratethemorphologyandstructureofinvitroreconstructedtissue-engineeredhumancornealepithelium(TE-HCEP)withseedercellsfromanuntransfectedHCEPcellline.·METHODS:TheTE-HCEPswerereconstructedinvitrowithseedercellsfromanuntransfectedHCEPcellline,andscaffoldcarriersofdenudedamnioticmembrane(dAM)inair-liquidinterfaceculturefor3,5,7and9days,respectively.Thespecimenswereexaminedwithhematoxylin-eosin(HE)stainingofparaffin-section,immunocytochemicalstaining,scanningandtransmissionelectronmicroscopy.·RESULTS:DuringinvitroreconstructionofTE-HCEP,HCEPcellsformeda3-4,6-7and8-10layersofanHCEP-likestructureondAMsinair-liquidinterfaceculturefor3,5and7days,respectively.Butthecellsdeceasedto5-6layersandthestructureofstraifiedepitheliumbecamelooseatday9.Andthecellsmaintainedpositiveexpressionofmarkerproteins(keratin3andkeratin12),cell-junctionproteins(zonulaoccludens-1,E-cadherin,connexin43andintegrinβ1)andmembranetransportproteinofNa+-K+ATPase.TheHCEPcellsinTE-HCEPwererichinmicrovillionapicalsurfaceandestablishednumerouscell-cellandcell-dAMjunctionsatday5.·CONCLUSION:ThemorphologyandstructureofthereconstructedTE-HCEPweresimilartothoseofHCEPinvivo.TheHCEPcellsinthereconstructedTE-HCEPmaintainedthepropertiesofHCEPcells,includingabilitiesofformingintercellularandcell-extracellularmatrixjunctionsandabilitiesofperformingmembranetransportation.TheuntransfectedHCEPcellsanddAMscouldpromisinglybeusedinreconstructionHCEPequivalentforclinicalcornealepitheliumtransplantation.