I believe generating hypotheses is a very worthwhile task for students because of the high cognitive demand it involves. They are constructing their own understanding because they are using their own ideas to come up with new hypotheses. According to Pittler, Hubbell, Kuhn, and Malenoski (2007), using these types of complex mental processes enhances students' understanding of content. I can think of no other way to put constructivist theory into practice than to have students complete these tasks, because students have to truly create their own ideas and understanding.
Testing hypotheses is a great example of constructionism because of the many ways we can implement this strategy. Of course, the most obvious examples occur in science where we conduct physical experiments to get results. However, this can go beyond the science classroom to incorporate other content areas. The advancement of technology has helped this instructional strategy become more effiicient because students can spend less time on gathering data, and more time interpretting it (Pittler, et al., 2007). This is important because one key idea of constructionism is that students build an artifact in order to share it with others so that they may learn and benefit from it. Allowing the builders of the artifact to spend more time interpretting data will allow for more thinking on analyzing the results and reporting them to others.
References:
Laureate Education, Inc. (Executive Producer). (2010). Program seven. Constructionist and constructivist learning theories. [Webcast]. Bridging learning theory, instruction, and technology. Baltimore, MD: Author.
Pitler, H., Hubbell, E., Kuhn, M., & Malenoski, K. (2007). Using technology with classroom instruction that works. Alexandria, VA: ASCD.