When something is not attended to by a person, even when it is right before them, they won't perceive it. This is known as inattentional blindness (Mack & Rock, 1998). Sometimes information missed due to inattentional blindness is trivial but inattentional blindness can become a problem when it hinders people from responding to something appropriately when a response is needed. When a visual cue is missed there can be an impact on decision-making. Variations in color luminance may also be a factor in one's ability to attend to something. For example, if a person is attending to a number of objects that are one color shade (for instance, dark green), it may be possible that this person might not see an additional object appear in their field of view (FOV) if it is the same color and shade. Conversely, the opposite might be true that a person is more likely to attend to the additional object if it is the same dark green color, opposed to an object that is colored a lighter green. This research investigated whether some variations of luminance of the same color (for example, dark green to light green) can affect one's ability to attend an additional object entering one's FOV. A scenario was presented to tank gunners that required them to observe objects of one color (dark green) while an additional object was briefly presented to them colored either dark green or light green. In this between-subjects study, 48 participants observed four dark green and four light green enemy tanks moving about the battlefield. Each was given a task that involved monitoring the dark green tanks only. During their monitoring, an additional vehicle (M981A3 FIST-V) briefly entered and exited their FOV. The additional vehicle was presented to 24 participants colored dark green. For the other 24, it was presented colored light green. This research addressed whether there was an association between color luminance, FOV, or focused attention and detection of the FIST-V. The results did not indicate an association between FOV and detection of the FIST-V [÷2(1, N = 48) = 0.08, p = 1.0]. Nor was there an association between focused attention and detection of the FIST-V using the following self-reporting questionnaires for determining levels of focused attention: ETAS [÷2(1, N = 48) = 2.06, p = 0.20], the CFQ [÷2(1, N = 48) = 0.75, p = 0.56], and the DAPI [÷2(1, N = 47) = 1.39, p = 0.75]. In the same manner, there was also no association between field dependence and detection of the FIST-V [÷2(1, N = 43) = 0.34, p = 0.75]. There was, however, an association between color luminance and detection of the FIST-V [÷2(1, N = 48) = 36.80, p < 1.0e-8]. / Master of Science

This study is an empirical analysis of the 2009 and 2010 forms of the Gyeonggi English Communicative Ability Test (GECAT) based on the responses of 2,307 students to the 2009 GECAT and 2,907 students to the 2010 GECAT. The GECAT is an English proficiency examination sponsored by the Gyeonggi Provincial Office of Education (GOE) in South Korea. This multiple-choice test has been administered annually at the end of each school year to high school students since 2004 as a measure of the students' ability to communicate in English. From 2004 until 2009, the test included 80 multiple-choice items, but in 2010, the length of the test was decreased to include only 50 items. The purpose of this study was to compare the psychometric properties of the 80-item 2009 form of the test with the psychometric properties of the shorter 50-item test using both Classical Test Theory item analysis statistics and parameter estimates obtained from 3-PL Item Response Theory. Cronbach's alpha coefficient for both forms was estimated to be .92 indicating that the overall reliability of the scores obtained from the two different test forms was essentially equivalent. For most of the six linguistic subdomains, the average classical item difficulty indexes were very similar across the two forms. The average of the classical item discrimination indexes were also quite similar for the 2009 80-item test and the 50-item 2010 test. However, 13 of the 2009 items and 3 of the 2010 had point biserial correlations with either negative or lower than acceptable positive values. A distracter analysis was conducted for each of these items with less than acceptable discriminating power as a basis to revise them. Total information functions of 6 subdomain tests (speaking, listening, reading, writing, vocabulary and grammar) showed that most of the test information functions of the 2009 GECAT were peaked at the ability level of around 0.9 < θ < 1.5, while those of the 2010 GECAT were peaked at the ability level of around 0.0 θ < 0.6. Recommendations for improving the GECAT and conducting future research are included.