Environmental Engineering Reference
In-Depth Information
illustrate pictorially the range of attitudes toward the use of modern Western
medicine from most receptive to most resistant to change in the community. A
correlation of the Guttman scale scores with such variables as age, education,
immigration status, and related value system variables identified population seg-
ments that would be most interested in educational material about alternative
medicating practices. This information provided a target group and suggested an
efficient use of limited educational resources while respecting the wishes of
those who were not interested in learning more about modern medicating prac-
tices. (See Pelto, 1970, for additional discussion of the Guttman scale.)
Likert scales provide an advance in reliability, validity, and ease of use
over the Guttman scale. A typical Likert survey scale is based on a 5-point
rating system. For example, concerning the topic “support and guidance in
securing employment,” excellent was 5, satisfactory was 3, and poor was 1 on
the scale. The Likert survey scale was easy to administer, and analysis con-
sisted of calculating the mode, mean, and range. The survey questions were
based on individual interviews and focus groups, grounding the questions in
the students' perception of the program. The survey provided some degree of
generalizability or representativeness concerning student opinions about the
program. The survey findings, in combination with interviews and observations,
made for credible and persuasive conclusions about the “consumer's view” of
the program.
A chi-square test provided an insight into enrollment trends in the study of
gifted and talented education programs. The test indicated that Latinos had the
most statistically significant increase in that program (Fetterman, 1988a).
Another popular nonparametric statistical tool in anthropology is the Fisher's
Exact Probability Test. However, all statistical formulas require that certain
assumptions be met before the formulas may be applied to any situation. A dis-
regard for these variables in the statistical equation is as dangerous as neglect
of comparable assumptions in the human equation in conducting ethnographic
fieldwork. Both errors result in distorted and misleading efforts at worst, and
waste valuable time at best.
Ethnographers use parametric statistics when they have large samples and
limited time and resources to conduct all the interviews. Survey and question-
naire work often requires sophisticated statistical tests of significance, such as
t tests. They are used to determine whether the means of two groups are sta-
tistically significant (Trochim, 2006b). For example, in a recent study of the
medical curriculum at Stanford University, we used a t test to determine if the
impact of an empowerment evaluation innovation significantly impacted cur-
ricular development. This innovation engages faculty and students in a collabo-
rative assessment and refinement of the curriculum (Fetterman & Wandersman,
2005). We compared student evaluations before and after the innovations to
Search WWH ::




Custom Search