https://padlet.com/charteredcollege/validityreliability Share your reflections en-us 2014-01-08 18:18:33 UTC 2022-08-21 10:47:21 UTC hello@padlet.com https://padlet.com/charteredcollege/validityreliability/wish/1858477644 I think we prioritise reliability over validity and that is something we could address. ]]> 2021-11-01 11:28:26 UTC https://padlet.com/charteredcollege/validityreliability/wish/1858477644 https://padlet.com/charteredcollege/validityreliability/wish/1952386235 Christine Counsel on Twitter once highlighted an end of year 9 history exam which had numerous short skills-based questions.
This meant (at least one could argue) that reliability was, therefore, higher than for some more "traditional" history exams which involved students writing extended answers for a smaller number of questions. 
Given the above discussion about reliability vs validity, it does make one think about the way my school has been doing KS3 assessments which model themselves after GCSE exams (albeit with numerous changes and alterations and mark schemes which don't simply use GCSE levels).
A summative end of year test which had more numerous short questions that still allowed students to demonstrate their skills and learning would (could?) be more reliable. 
However, the skill of a historian is demonstrated in the ability to argue in written prose, so while a student could show their appreciation of chronology or causation in a variety of short answer questions, these are not discrete skills, but only part of the larger skillset demanded of a historian.
This links back to the larger problem which historians have, which is exploring what we are wanting students to achieve and what aspects of being a historian do we want them to master at any given stage - these are issues which need to be addressed before thinking about reliability or validity. 
So, in sum - more numerous short answer questions => increased reliability, but lower validity (less representative of "true" history), while fewer longer-answer questions=> lower reliability, but higher validity as more representative of "true" history. ]]> 2021-12-16 18:15:06 UTC https://padlet.com/charteredcollege/validityreliability/wish/1952386235 https://padlet.com/charteredcollege/validityreliability/wish/1953467062 Not sure that I agree that there is always a trade off between these. Surely you can have a test on one science topic, increase it's length to add, say, a 6 mark written Q. This could still be marked in such a way to give reliable data while also making it more valid??
 ]]> 2021-12-17 10:17:38 UTC https://padlet.com/charteredcollege/validityreliability/wish/1953467062 https://padlet.com/charteredcollege/validityreliability/wish/1954700592 I think the existing responses are very interesting. In my school, I feel that we have prioritised validity over reliability. For example, when we ensure all children can access the assessment, we introduce more variables. I think we should move towards a slightly more reliable model in order to get a better balance. It’s important to take several assessments and your knowledge of the pupil to triangulate judgements.

Daisy Christodolou’s book, ‘Making good progress’ is an interesting read. I’d highly recommend it.]]> 2021-12-18 10:29:31 UTC https://padlet.com/charteredcollege/validityreliability/wish/1954700592 https://padlet.com/charteredcollege/validityreliability/wish/2048272492 There has to be a trade off between validity and reliability.  As stated to make an assessment more reliable it needs to be increased in length.  This however brings it’s own set of issues, assessments preferably need to be done within one lesson so not too much teaching time is lost and to help ensure that students don’t have the opportunity to find out what’s on the assessment before taking it.  Another consideration with making assessments longer is that they may unintentionally disadvantage lower attaining students whom do not have the ability to concentrate for extended periods of time.  One solution that we use is that shorter tests are used at the end of units of work thereby narrowing the test.  The issue with this is because a broad range of questions aren’t used validity may be reduced.  Another key area to consider when thinking about reliability is moderation.  Whilst this is put in place when marking mock exams or course work it may not be for shorter end of unit tests due to not wanting to increase teacher workload.

]]> 2022-02-15 10:30:34 UTC https://padlet.com/charteredcollege/validityreliability/wish/2048272492