In inquiry learning in science education, learners are involved in scientific activities like searching literature, formulating hypotheses, gathering and interpreting scientific data together. These tasks are considered highly challenging and even overwhelming if not supported appropriately with scaffolding, small group scripting or expert commentaries (e.g., Linn, 2006; Kollar, Fischer & Slotta, 2007). Moreover, when facing the problems in conducting typical inquiries such as hypothesis formation or data interpretation the students often refrain from seeking help from either their peer learners or their teacher (van Joolingen, de Jong, Lazonder, Savelsbergh, & Manlove, 2005). Better help seekers, however, might benefit substantially more from learning and collaborating in these challenging environments. Research findings in other areas of help seeking research suggest that learners differ substantially in their help seeking behaviour and better help seekers do learn more, but help seeking should be facilitated by instruction and the resulting patterns of classroom interaction (Aleven, Stahl, Schworm, Fischer, & Wallace, 2003). In our study, we focus on classroom scripts as specific aspects of the instructional approach (Seidel, Rimmele, & Prenzel, 2005). Research on the question how different patterns of classroom interaction for collaborative inquiry learning support or hinder the help-seeking processes is rare so far. From the perspective of help seeking research, not asking for help when it is needed is not a problem which is exclusive to collaborative inquiry learning classrooms; help-seeking research has indicated that this phenomenon is generally widespread in a variety of different educational settings (Aleven et al., 2003; Newman, 2000; Ryan, Pintrich, & Midgley, 2001). Although there is indication that the instructional approach can influence help seeking (Karabenick & Newman, in press), empirical investigations on the effects of different instructional approaches are rare. As well there is less research done on how different instructional approaches resulting in different classroom interaction patterns, in turn, may affect on boys and girls when collaborating in mixed- or single-gender computer-supported groups (e.g., Scanlon, 2000; Underwood, Underwood, & Wood, 2000). Methodologically, there has been a focus on questionnaire-based research in the area of help seeking and there is a clear need for empirical studies including measures of help seeking behaviour and consequential learning outcomes in real learning contexts. This study examined how the structure of the classroom script (low vs. high) and computer-supported collaboration in mixed- vs. single-gender pairs at the secondary school science education affect on help-seeking processes and learning outcomes.

A total of 108 secondary school students worked in pairs (n = 54 pairs; low-structured 35 vs. high-structured 19; mixed-gender = 21; single-gender = 33) on a physics project “How far does light go?” using a Web-based Inquiry Science Environment (Slotta & Linn, 2000). In the high-structured classroom script condition, teacher-led plenary activities following each inquiry step done in pairs. In the low-structured classroom script condition, students worked in pairs without the teacher-led plenary phases. Using the sampling method (in total 20 min/each pair) in an analysis of 54 screen-capture and audio videos, help-seeking events were identified and coded in more detail by applying Nelson-Le Gall’s (1981) model (Cohen’s Kappa 0.69). Pre- and post-tests of students’ knowledge of physics were taken using six multiple-choice items and one item with a free answering format and an established coding procedure (Bell, 2004; the agreement rate 86 % - 99 % across schemes).

The results indicate that the amount of help sought was lower in the high-structured classroom script condition than in the low-structured classroom script condition (F = 6.422, p < .05, η2 = 0.118). Students asked more executive help even with the increasing amount of help requested in the low-structured classroom script condition (F = 6.846, p < .05; η2 =0.125). There were hardly any differences on the student help-seeking behaviour and learning outcomes in mixed-gender pairs and single-gender pairs. Students in both conditions acquired knowledge, but there were no differences in high- and low-structured classroom script condition. Based on the results, it seems that the differently structured classroom conditions did not enhance beneficial help seeking processes with respect to learning, such as higher quality or more specific help requested.