谁能建议一个好的测试工作估算模板
我正在寻找测试工作量估算模板。如果有人有好的模板请与我分享。
I am looking for a test effort estimation template. Please share with me if anyone has a good template.
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（3）
没有模板，因为估算过程必须根据变化进行调整。有不同的估计方法，考虑过程的不同部分。
例如，您可以采用基于因素的方法，并假设测试将花费 1.5 倍的开发工作量。这是一种狭隘的方法，因为并非所有测试工作量都等于 1.5 倍的开发工作量。
其他方法包括进行功能点分析以及估计或估计每天可以编写多少个测试用例以及每天可以执行多少个测试用例。测试时间中是否包括数据设置？
另一种方法考虑了可用资源、环境的稳定性、每种资源的技能以及它们对项目的贡献程度。
在所有这一切中，您需要考虑正在测试的内容；用户界面，客户端-服务器系统；它包括分布式服务吗？你在测试它们吗？您只是关注功能测试还是需要进行一些安全测试以及负载/性能和压力测试？所有这些都可能需要相当长的时间来执行。
估计应该由经验丰富的测试人员完成，该测试人员以前曾在系统类型上工作过，最好是在组织内工作过。
因此，没有一个模板，也没有一堆变量需要填写才能得出准确的时间数。每次测试所花费的时间都不同。如果有缺陷，则需要包括返工和重新测试时间。你会提出多少缺陷？这不是一个你可以回答的问题，但如果史蒂夫天真地向他的测试人员提供估计，他还应该指出其中仍然存在多少错误。
关于估算要记住的基本一点是它只是一个估算。
There is no template because the estimation process has to be tailored to change. There are different approaches to estimation which consider different parts of the process.
For instance you can take a factor based approach and assume that testing will take 1.5 times development effort. This is a narrow-minded approach as not all testing effort is equal to 1.5 development effort.
Other approaches include doing function point analysis and estimating or estimating how many test cases can be written per day and how many can be executed per day. Do you include data setup in the testing time?
Another approach factors in the resources available, the stability of the environment, the skill of each resource and how much they can contribute to the project.
In all of this you need to consider what is being tested; a ui, a client-server system; does it include distributed services? Are you testing them? Are you just looking at functional testing or do you need to do some measure of security testing as well as load/performance and stress testing? All of which can take considerable time to execute.
Estimation should be done by an experienced tester who has worked on the types of system and preferably within the organisation before.
So there is no one template and nor is there a bunch of variables to fill out to come out with an accurate number of time. Testing takes different amounts of time, each and every time you do it. If you have defects then the rework and retest time needs to be included. How many defects will you raise? Not a question you can answer but if Steve is naively providing estimations to his testers he should also indicate how many bugs are still in there.
The fundamental thing to remember with estimation is that it is an estimate.
估计工作量是开发和测试中最困难的事情之一。在我们公司，当开发人员向测试小组提交更新（包含正式文档）时，开发人员有责任估计所需的测试量。我不明白事情会怎样。
我对测试的估计总是有点猜测，但一般来说，如果我开发一些东西需要一周，那么我会估计需要 2 天的测试。显然，这不需要在所有示例中都很好地工作，并且需要根据所讨论的变化进行调整。例如，我可能会花一周的时间来开发可以很快进行测试的更新。
Estimating effort is one of the hardest things to do in development and testing. In our company when a developer submits an update (complete with the formal documentation) to the Testing gang, it is the responsability of the developer to estimate the amount of testing required. I do not see how that could be otherwise.
The estimates that I give to Testing are always a bit of a guess, but in general if I develop something for a week then I would estimate 2 days testing. Obviously, this does not need work well in a ll examples and need to be adapted based on the change in question. For example, I might spend a week working on an update that can be tested very quickly.
估计考试时间会让你的头爆炸。抓住一些地牢骰子并滚动它们几乎更容易（也更准确）
开发人员可能会花费 1 小时进行需要数天测试的修复，因为该修复对于系统来说是如此不可或缺，以至于必须测试它的所有其他方面。
开发人员还可以花 2 天的时间在网页上修复 CSS、标记等，而在多个浏览器中只需要 4 个小时的测试。
开发人员可以定义“应该测试”的内容，但请不要依赖它，因为我们很自豪，并且在我们看来（有时是我的想法）人们应该按照程序的预期方式使用程序。为什么它应该允许用户名是 Chuck^Norris2010。该用户需要被打……我永远不会测试这样的用户名，我永远不会想到测试它。
测试的因素有很多，但我知道大多数开发人员错过的一项测试是边界测试和流程测试。这些类型的测试很难衡量。
您知道需要测试什么。随着时间的推移，您不断测试软件，您将学习如何更好地估计测试，因为您将了解系统。
开发人员将能够告诉您更改了什么，以及他们知道需要测试什么。
估计不是马上就能得到的事情。它需要经验和知识，以及领域知识。
当您进行测试时，您将获得时间，并且当您记录测试项目各个方面所需的时间时，您将能够更好地给出更好的估计。
Estimating test time will cause your head to explode. It's almost easier (and more accurate) grabbing a few dungeon dice and rolling them
A developer could spend 1 hour doing a fix that requires days of testing because that one fix was so integral to the system that all other aspects of it have to be tested.
A developer could also spend 2 days fixing CSS, markup, etc. on a webpage that only requires 4 hours of testing in multiple browsers.
Developers are ok in defining what "Should be tested" but please don't rely on that, because we're prideful and in our minds (well my mind at times) people should just use the program the way it's meant to be used. Why should it allow a username to be Chuck^Norris2010. that user needs to be slapped... I would never test a username like that, I would never think of testing that.
There are many factors of testing, but the one test I know most dev's miss is boundary testing, and flow tests. These types of tests are hard to measure.
You know what needs to be tested. And as time goes on and you keep testing software you will learn how to better estimate the testing because you will know the system.
A developer will be able to tell you what was changed, and what they know needs to be tested.
Estimating isn't something that comes right away. It takes experience and knowledge, and field knowledge.
As you test you'll derive times, and when you record how long it takes you to test each aspect of a project, you'll be better equipped to give better estimates.