The authors investigated the effects of differing question-stem designs (detailed stem, simple stem, and no stem) on elementary students regarding the quality of the posed questions by students, as well as their science reports, science achievements, and critical thinking ability. A quasi-experimental design was adopted. One hundred fifth-grade Taiwanese students were assigned to three groups: detailed stem (G1), simple stem (G2), and no stem (G3). Three pupils formed a team, with each group having 11 teams. The members of each team needed to collaborate for proceeding with an online questioning activity. The results showed that the detailed-stem group had a better questioning quality. For science reports, the detailed-stem group and simple-stem group had a significantly better effect than the no-stem group. Regarding science achievement, there was no significant difference among the three groups. And finally, for critical thinking, the detailed-stem group and no-stem group had a significantly better effect than the simple-stem group.