Abstract:
The purpose of this study is to describe the validity and reliability of science process skills based
test instruments and the quality of science process skills based on test instruments is viewed from
the level of difficulty, power difference, and deception in material objects and their properties for
Grade V SDN 010 Tarakan. The type of research used in this research is Formative Evaluation
Research and Development design. Research Results there are 30 items in the form of multiple
choice, the initial steps are validation by a team of experts in a very feasible category. The one to
one trial was conducted in conjunction with the validation of the one-to-one trial experts to get
90% of students' responses in the very effective category. The results of the SPSS 21 program
analysis show 20 questions in the valid category, the reliability is r11 = 0.585> r table = 0.284.
The results of the analysis of the AnatesV4 program show the level of difficulty with difficult
categories of 1 item problem (3.3%), the level of difficulty of the questions that are in the easy
category a number of 1 items (3.3%), the level of difficulty of the questions which are in the
category of very easy number 2 items of questions (6.7%), the level of difficulty of the questions in
the moderate category, 26 items of questions (86.7%), the results of the average power difference
with a good criterion of 12 questions (40%), and an effective deceiver as many as 28 items
(93.3%).
