You are here: Home / Publications / IPN-Blätter / Factors for Independent Experimentation

Factors for Independent Experimentation

An article by Lina Boyer - winner of the 2015 IPN Prize for outstanding empirical master thesis

experiment_boyer.png
Figure 1: Structure of the experiment for uniformly accelerated motion
Experiments have a key function when it comes to scientific knowledge acquisition. Among other things, they help examine scientific theories and generate knowledge. It turns out, however, that students often have problems approaching experiments systematically and strategically. The majority of students is unable to independently experiment without guidance. To best target support it is necessary to know what personal factors are crucial for independent experimentation. In recent studies factors such as prior knowledge, interest or language skills were identified as prerequisites for successful experimentation. Most studies, however, do not distinguish between individual phases of experimentation although one can assume that the individual factors are not equally relevant to all phases of experimentation. For example, it is expected that a person’s thematic knowledge is more important for hypothesizing than for performing measurements. Therefore, the research question of this study was: What personal factors are important in each phase of the experiment and for successful experimentation overall?

The linguistic skills of test subjects were ascertained to clarify the impact of potential personal factors, knowledge, and interest. Students were then asked to plan, carry out, evaluate and discuss ten open experiments. All ten experiments pertained to mechanics on a secondary school level. Additionally, a protocol should be written for each of these experiments. These protocols are the basis for assessing the success of an experiment. The different phases of experimentation planning, implementation, monitoring, evaluation and discussion were evaluated. The sum of all phases determines the success of the whole process.

Multiple regression analyses were calculated for each stage of experimentation and for the entire process. There were no significant regression models for implementation and monitoring. The standardized regression coefficients for the other phases are listed in Table 1 below.

Table 1: Standardized regression coefficients of the experimental success of the entire process and the various stages of experimentation

The results indicate that several factors influence the experiment. The study also provides evidence that these factors are weighted differently in the phases. Interest seems to be a critical factor when it comes to designing experiments. Physics class should therefore promote interest, so students can experiment successfully. The necessary expertise as well as the language skills are relevant for the evaluation of experiments. People with better language skills evaluate their experiments more successfully than those with poor language skills, even if they have the same amount of knowledge. This may also result in this effect because the experimental protocols were evaluated as a measure of success. One possible consequence for example could be to consciously practice the linguistic design of experimental protocols.

Additional factors relevant for the individual phases of experimentation should now be integrated into the study to obtain a more differentiated view of the support options for experimentation. Various experimental strategies or manual skills for example could be taken into account. If the strengths and weaknesses of a student are known, individually tailored support could be offered in the various phases of experimentation, thereby promoting their experimental skills.