In choosing an assessment, validity should have priority over reliability for two reasons.
First, validity is arguably more important than reliability. When you select an assessment, you want it to measure the thing that it is supposed to measure. You do not want to select, for example, a reading comprehension assessment that actually measures only the student’s ability to decode words. Validity is the measure of how well your assessment actually measures what it is supposed to measure.
Second, and perhaps more importantly, any assessment that is valid will also be reliable. That means that, by picking a valid assessment, you will also get a reliable assessment. An assessment is reliable if it returns more or less the same score every time someone with the same ability level takes it. If a test is valid, it cannot help but be reliable. If it is truly measuring what it should measure, it has to return more or less the same result each time a person with the same ability level takes it.
For these reasons, it makes more sense to seek validity over reliability.