溫馨提示：需求數(shù)量不同，價(jià)格不同。請(qǐng)聯(lián)系我們，確認(rèn)當(dāng)前新的報(bào)價(jià)！
flexMIRT是可用的先進(jìn)的IRT軟件，一個(gè)用于項(xiàng)目分析和測(cè)試評(píng)分的多層次、多維和多組項(xiàng)目反應(yīng)理論（IRT）軟件包。flexMIRT適用于各種一維和多維項(xiàng)目反應(yīng)理論模型（也稱(chēng)為項(xiàng)目因素分析模型）以及診斷分類(lèi)模型，適用于任一數(shù)量組的單水平和多水平數(shù)據(jù)。
flexMIRT是易于使用的
基于Windows的flexMIRT具有圖形用戶(hù)界面（GUI），有64位版本。它具有直觀的語(yǔ)法，可以在命令行模式下無(wú)縫運(yùn)行，以實(shí)現(xiàn)大批量生產(chǎn)。用戶(hù)手冊(cè)演示了各種帶有注釋的輸入和輸出的分析。
flexMIRT是靈活的
我們的IRT軟件可以適用于各種項(xiàng)目級(jí)模型，包括1PL、2PL、3PL、分級(jí)響應(yīng)模型、廣義部分信用模型和診斷分類(lèi)模型。它能夠適應(yīng)多個(gè)維度、數(shù)據(jù)內(nèi)的依賴(lài)性（即多水平）、交叉隨機(jī)效應(yīng)、協(xié)變量和潛在特征的偏離正態(tài)性。
flexMIRT是快速的
隨著現(xiàn)代CPU架構(gòu)朝著多核設(shè)計(jì)的方向發(fā)展，flexMIRT使用并行處理，通過(guò)將負(fù)載自動(dòng)分配到多個(gè)可用的核心或處理單元來(lái)進(jìn)一步加快計(jì)算速度。在可能的情況下，flexMIRT使用降維來(lái)自動(dòng)減少多維或多級(jí)模型的集成維數(shù)。
flexMIRT可以處理大型數(shù)據(jù)集
一種新設(shè)計(jì)的內(nèi)存分配方案有助于flexMIRT有效地處理數(shù)千個(gè)項(xiàng)目和數(shù)百萬(wàn)受訪者。適合需要物品反應(yīng)理論模型的任一尺寸分析。
支持的版本
穩(wěn)定版本：版本 3.6.5
候選版本：不適用
已棄用版本：版本 3.5x
所有支持的版本：版本 3.0+
注意：
目前，目前不提供任一程序的本機(jī)Mac版本。但是，如果您有雙啟動(dòng)設(shè)置或 Windows 模擬器，只要您的硬件由于已知的兼容性問(wèn)題而不使用 Apple 專(zhuān)有的基于 ARM 的芯片，任一程序都應(yīng)該正常運(yùn)行。請(qǐng)注意，只能提供對(duì)使用該軟件的支持，無(wú)法提供有關(guān)設(shè)置 Mac 以允許程序可用的指導(dǎo)。Parallels Desktop 是常用的模擬器。VirtualBox是Oracle的免費(fèi)替代品，但與Parallels相比，它可能有點(diǎn)慢。
英文：
A multilevel, multidimensional, and multiple group item response theory (IRT) software package for item analysis and test scoring. flexMIRT fits a variety of unidimensional and multidimensional item response theory models (also known as item factor analysis models) as well as diagnostic classification models to single-level and multilevel data in any number of groups.
flexMIRT is easy to use
Windows-based flexMIRT has a graphical user interface (GUI) and is available in 64-bit version. It has an intuitive syntax and can be run seamlessly in a command line mode for high volume production. The user’s manual demonstrates a variety of analyses with annotated input and output.
flexMIRT is flexible
Our IRT software can fit a wide variety of item-level models including 1PL, 2PL, 3PL, Graded Response Model, Generalized Partial Credit Model, and Diagnostic Classification Models. It is able to accommodate multiple dimensions, dependence within the data (i.e., multilevel), crossed random effects, covariates, and departures from normality in the latent trait(s).
flexMIRT is fast
As modern CPU architecture trends toward multi-core design, flexMIRT uses parallel processing to further speed up computations by spreading the load automatically to the multiple available cores or processing units. When possible, flexMIRT uses dimension reduction to automatically reduce the number of dimensions of integration for multidimensional or multilevel models.
flexMIRT can handle large datasets
A newly-designed memory allocation scheme helps flexMIRT efficiently handle thousands of items and millions of respondents. Great for any size analysis where item response theory models are needed.
Supported Versions-flexMIRT
Stable release: Version 3.6.5
Release candidate: n/a
Deprecated versions: Versions 3.5x
All supported versions: Versions 3.0+
Note:
Currently we offer no native Mac version of any of our programs currently available. However, if you have a dual-boot set-up or a Windows emulator, any of our programs should run without issues as long as your hardware is not using Apple’s proprietary ARM-based chips because of known compatibility issues. Note that we can only provide support for using the software and are unable to provide guidance on setting up a Mac to allow for a program to be usable. Parallels Desktop is the most frequently used of the emulators. VirtualBox is a free alternative from Oracle, but it can be very slow compared to Parallels.
