Summative assessments are tricky. When building an assessment, we can't just put together some questions that deal with the same material taught; much more goes into building a summative assessment than just throwing together some questions and hoping for the best. I have made an example assessment, and rather than include a document here, I made this assessment online to make it more secure. It is password protected, the order of the questions is randomized and the answer choices are randomized- all of this to help with the security of the exam. Now, since I made it specifically for this course, it is not a real assessment that will be given. For this reason, I can tell you that the password is iloveSpanish (all one work, capital S) so that you can access it and take a go with your knowledge on the Spanish conditional tense and vocabulary from a unit entitled "De Viaje". Click here to access the test.
As in any summative assessment, one must check for validity, reliability, and security. I have given you the information on the security of the document. I didn't use any test that I would give in the classroom- either online or face to face- because that would defeat the purpose of attempting to have a secure test, right? Even if the test was password protected and randomized, once information is on the web, it's always on the web. And people that know me might be able to find it... even though it seems impossible, I would not compromise test security in that way. When looking at the validity of a summative assessment, we must consider if the material on the test matches what was taught in the course content and is aligned with the state and national standards for that course. We also need to consider if enough questions assessing the same topic were given to ensure that the grade will represent what the student knows and not what they do not know. This can be tricky, but consider giving a vocabulary test over 6 words when you taught 54. How fair would that be to base a student's summative grade on such a small percentage of the material? When it comes to reliability, we have to make sure our questions are fair. This means that we need to test it over time, as well as look at the test data that shows statistics about questions that are often missed. This will help us determine the reliability of the test. It also answers helps answer the questions about if the course content sufficiently covers the material in the assessment. The assessment that is included has not been tested for reliability because no one has ever taken the test. I can already say that the validity is low due to lack of questions. There really aren't enough questions on the test to consider it valid because it is over grammar as well as vocabulary; how can 10 questions accurately measure student knowledge over that much content?!
0 Comments
Leave a Reply. |