Editing Thematic Apperception Test (section)

===Reliability===
[[Internal consistency]], a reliability estimate focusing on how highly test items correlate to each other, is often quite low for TAT scoring systems. Some authors have argued that internal consistency measures do not apply to the TAT. In contrast to traditional test items, which should all measure the same construct and be correlated to each other, each TAT card represents a different situation and should yield highly different response themes.<ref name="problems"/> Lilienfeld and colleagues <ref name="lilienfeld" /> countered this point by questioning the practice of compiling TAT responses to form scores. Both [[inter-rater reliability]] (the degree to which different raters score TAT responses the same) and [[test–retest reliability]] (the degree to which individuals receive the same scores over time) are highly variable across scoring techniques.<ref name="murstein" /> However, Murray asserted that TAT answers are highly related to internal states such that high test-retest reliability should not be expected.<ref name="lilienfeld" /> Gruber and Kreuzpointner (2013) developed a new method for calculating internal consistency using categories instead of pictures. As they demonstrated in a mathematical proof, their method provides a better fit for the underlying construction principles of TAT, and also achieved adequate [[Cronbach's alpha]] scores up to .84 <ref>Gruber, N. & Kreuzpointner, L. (2013). Measuring the reliability of picture story exercises like the TAT. PLoS One, 8(11), e79450. doi:10.1371/journal.pone.0079450</ref>