Category: Writing

Identify the threat to external validity in the following
examples:
a. A teacher read research reporting great results from
using a particular remedial reading program. She tried the program in her
classroom but did not get the same results.
b. A teacher read about a new method of instruction that was
very successful with regular classes. The method did not work when she tried it
with her special education students.
c. A research journal reported that a new antismoking fi lm
had brought about a signifi cant change in students’ attitudes. After viewing
the fi lm, students expressed many more negative attitudes toward smoking than
they had on a prefi lm measure. A teacher decided to use the fi lm with his
middle school students but did not want to take time to administer the
“pretest.” He administered an attitude scale following the fi lm but did not fi
nd marked negative attitudes toward smoking.

Identify the threats to internal validity in each of the
following experimental studies:
a. At the beginning of the school year, an elementary
teacher set up reading stations in her room with supplementary reading
materials in each. At the end of the year, she reported that the interest of
her class in reading was higher than in any of her previous classes. She
recommended that all teachers set up reading stations.
b. A researcher wanted to investigate the effectiveness of
new teaching materials designed to increase students’ verbal skills. One class
used the new materials, and another class used conventional workbooks. Both
classes were given a test of verbal skills at the end of the semester.
c. A study was conducted in three thirdgrade classrooms
investigating the effect of three methods of teaching math. The teacher could
choose the method she wanted to use but could use only one method with her
students. At the end of the semester, all the third-graders were given the same
math achievement test.

Evaluate the following research designs with respect to
methods used and the control provided. Make suggestions for improvements if
needed.
a. A researcher wanted to ascertain if homogeneous grouping
improves learning in a fi rst course in biology. The researcher designated one
of two high schools in a small city to serve as the experimental school and the
other as the control. Both schools had approximately the same number of
students in each of four sections of science. In the experimental school,
pupils were grouped homogeneously on the basis of aptitude and scores on
achievement tests in science. In the control school, pupils were placed in
sections at random. At the end of the year, all pupils were given a
standardized test in biology. Statistical tests showed the experimental group
to be superior on the test. The researcher concluded that homogeneous grouping
results in greater learning in biology.
b. A history teacher was concerned about her students’ lack
of knowledge of their state and national governments and of current events. She
decided to experiment with some new materials and methods to determine if she
could obtain improvement. In classes A and B, she introduced the new materials
and methods. In classes C and D, she used the traditional methods. Classes A
and B were administered both the pretest and the posttest; classes C and D were
administered only the posttest. When comparisons were made on the posttest, classes
A and B were found to be superior. Their superior performance was attributed to
the new materials and methods.

1. You have the following technical information from two
tests: On which test would a student’s score be expected to fl uctuate the most
on repeated administrations?
2. What are the sources of error that affect the reliability
of a test? Give an example of each.

A 100-item test was split into two halves, and the
split-half coeffi cient of correlation was found to be .60. Calculate the
reliability coeffi cient for the full-length test.

1. Determine the standard error of measurement for a test
with a standard deviation of 16 and a reliability coeffi cient of rxx= .84. How
would you interpret this standard error of measurement?
2. Check the test manual for the achievement test being used
in your school. What type of reliability data are reported there?

1. Explain how a mathematics achievement test could be
judged to have high validity in one mathematics class and low validity in
another mathematics class.
2. Criticize the following statement: The reliability of the
intelligence test is .90. Therefore, you can assume that the test scores can be
interpreted as measuring intelligence

1. List and briefl y describe the instruments available for
recording data in observational research.
2. What type of instrument would be most appropriate to
measure each of the following?
a. To determine if high school chemistry students can use
laboratory scales to weigh specifi ed amounts of a given chemical compound
b. How students in the various elementary schools in Brown
County compare in math skills
c. How parents feel about an extended school day for
elementary schools in the district
d. The general verbal and nonverbal abilities of a child
with dyslexia
e. To study bullying in an elementary classroom
f. To get a major professor’s evaluation of the potential of
a student for advanced work in chemistry
g. To get a quick measure of students’ attitudes toward the
extracurricular programs available at the school

1. Interpret this statement: “A statistical relationship
between two variables is a necessary but not a suffi cient condition for
inferring a causal relationship between the variables.”
2. What alternative hypotheses would you want to explore if
your causal-comparative research indicates that there is a relationship between
self-concept and achievement in school?