The Top Ten Myths and Urban Legends about Likert scales
Ten Common Misunderstandings, Misconceptions, Persistent Myths and Urban Legends about Likert Scales and Likert Response Formats and their Antidotes
James Carifio and Rocco J. Perla. Journal of Social Sciences 3 (3): 106-116, 2007
Myth 1—There is no need to distinguish between a scale and response format; they are basically the same “thing” and what is true about one is true about the other.
Myth 2—Scale items are independent and autonomous with no underlying conceptual, logical or empirical structure that brings them together and synthesizes them.
Myth 3—Likert scales imply Likert response formats and vice versa as they are isomorphic.
Myth 4—Likert scales cannot be differentiated into macro and micro conceptual structures.
Myth 5—Likert scale items should be analyzed separately.
Myth 6—Because Likert scales are ordinal-level scales, only non-parametric statistical tests should be used with them.
Myth 7—Likert scales are empirical and mathematical tools with no underlying and deep meaning and structure.
Myth 8—Likert response formats can without impunity be detached from the Liker Scale and its underlying conceptual and logical structure.
Myth 9—The Likert response format is not a system or process for capturing and coding information the stimulus questions elicit about the underlying construct being measured.
Myth 10—Little care, knowledge, insight and understanding is needed to construct or use a Likert scale.
Artikel komplit dapat didonwload disini