If I assign participants to a condition (between-subject) by the order they join the research (e.g., 1st participant goes to cond. A, 2nd part. to cond. B, 3rd part. to cond. C, 4th part. to cond. A, 5th part. to cond. B, 6th part. to cond. C, 7th part. to cond. A, and so on...).
What kind of design is that? Is it possible to call it "Blocked Randomization"? Any studies that used a similar approach?
PS: I am recruiting participants online and once they agree to participate, they'll be assigned to a condition depending by who's turn (e.g., I started by assigning part. 1 to cond. A, it follows that part. 2 goes to cond. B, 3 to C, 4 to A, and so on...)
Thanks in advance
You're not really employing randomization if you assign them to conditions based on the response order. Thus, you're not really doing an experiment in the traditional sense. Also recognize that early respondents and late responders are significantly different from one another (see Armstrong & Overton, 1977 for old-school, but still relevant, evidence), meaning you have a real chance of biasing your results based on the assignment method. This could single-handedly red flag your investigation.
Instead, why don't you randomly assign respondents to treatments and then model response time as a post-hoc test? This would maintain the integrity of your experiment. If you use an online experiment (e.g., Qualtrics/QuestionPro), you can also build in random assignment and equalize group assignment between treatments, so randomly assignment isn't that hard to achieve, and you don't have to constantly monitor your participants when the software does it for you.
If that's not possible, then you should probably include measures (controls) of proactivity and maybe altruism in your study since these sometimes bias results. This enables you to control for selection effects somewhat.