**Thesis Type:** Postgraduate

**Institution Of The Thesis:** Middle East Technical University, Turkey

**Approval Date:** 2006

**Thesis Language:** English

**Student:** Hüseyin Delil

**Supervisor: **BEHİYE UBUZ

The purpose of this study was to analyze geometry problems in a widely used sixth-, seventh-, and eighth-grade Turkish elementary mathematics textbooks series based on the cognitive assessment frameworks of the most recent TIMSS, the Trends in International Mathematics and Science Study 2003. To compare geometry problems in the textbooks and TIMSS 1999, in which Turkish students poorly performed, the cognitive behaviors that the problems required were determined and categorized. After the analysis, it was found that the two most frequent behaviors that the problems require are computing and applying with a total percentage of 72, in case of the textbooks. In TIMSS 1999 geometry problems case, however, applying and analyzing are the most frequent cognitive behaviors with a total percentage of 47. It was also found that a great majority of 22 behaviors of the framework were either not represented or underrepresented by the textbooks geometry problems. When we consider the four major categories of behaviors, 86 percent of the textbooks geometry problems required behaviors belong to two cognitive domains: Knowing Facts and Procedures or Solving Routine Problems. TIMSS 1999 geometry problems, however, mostly belong to Solving Routine Problems or Reasoning with a percentage of 65. In both the textbooks and TIMSS 1999 cases, a relatively small part of the problems required behaviors belong to Using Concepts. The results are discussed in the light of Turkey’s performance in TIMSS 1999 and some suggestions related to the textbook problems were given.