Scientific Attitude Survey Paper
Scientific Attitude Survey Paper
Scientific Attitude Survey Paper
Background
Motivation of students to be interested in science, their attitudes toward science, their views
of scientists, and their desire to become scientists have been investigated by science educators
for many years. Interest has grown since Nolls 1935 investigation of the measurement of scientific attitudes, and Mead and Metrauxs 1957 investigation of high school students image of
scientists. More recently, science educators have been struggling with defining science attitudes
(Shrigley, Koballa, & Simpson, 1988) and differentiating among attitudes, beliefs, and values
(Koballa, 1988). A critical review of science attitude measurement by Munby (1983) included
30 studies that used the Scientific Attitude Inventory (SAI). Munby reported that the SAI was
the most popular instrument of its type at that time, as he raised questions about its validity. The
purpose of this article is to report the development, field test, and availability of a revision of
the original SAI, the SAI II.
Requests for permission to use the SAI and/or this revision have been received recently
from Boston University to study attitude development within a National Science Foundation
(NSF) grant, from The University of Southern Mississippi to study attitude development in high
Correspondence to: R.W. Moore
1997 by the National Association for Research in Science Teaching
Published by John Wiley & Sons, Inc.
CCC 0022-4308/97/040327-10
328
school biology classes, from Sheffield Hallam University for a study of attitudes of first-year
students of primary education toward science, from several students at Temple University studying attitude development among high school students, and from the State University of New
York at Brockport for a study of the effect of a businessschool partnership on students attitudes toward science. Dimensions of the attitudes toward science addressed by the SAI are at
the forefront of interest in the development of a literate citizenry. For example, the knowledge
or attitudes that scientific ideas are subject to change, that science cannot answer all questions,
and that scientific work requires creativity are part of the subject of this instrument and are identified as important dimensions of the nature of science by Project 2061 (American Association
for the Advancement of Science [AAAS], 1990).
The Original SAI
The development of the SAI was reported in the Journal of Research in Science Teaching
over 25 years ago (Moore & Sutman, 1970). The SAI has been used extensively throughout the
world (Baker, 1985; Finson & Enochs, 1987; LaShier & Nieft, 1975; Welch, 1972). The authors
are aware that it has been translated into Hebrew and Thai. A request for permission to translate it into Spanish was received, but there is no evidence of that or any additional translation.
In developing the instrument initially, the plan was to develop position statements representing the universe of content, namely, attitudes toward science. Twelve position statements,
six opposing positive and negative statements, were developed. The position statements were
intended to represent both intellectual and emotional attitudes. Then, a pool of attitude statements was developed for each of the position statements. The attitude items were submitted to
a panel of judges who judged each attitude statement with respect to whether it represented a
particular position statement. The panel of judges consisted of 4 science educators, 4 practicing
scientists, and 2 liberal arts science professors. On the basis of the judges judgments, attitude
statements were selected from the pool for use in the instrument. Content validity for the SAI
was claimed on the basis of the judgments of the panel of judges (Kerlinger, 1986; Gable, 1993).
The SAI consists of 60 Likert-type items. There are 5 attitude statements for each of 12 position statements, 6 opposing positive and negative statements. Each set of five statements represents a particular position statement and is referred to as a scale. Half of the position statements are positive and half are negative; thus, half of the 60 attitude statements are positive and
half are negative. The SAI was field tested with three classes of 10th-grade biology students.
The field test is reported in the Journal of Research in Science Teaching (Moore & Sutman,
1970). Scoring the original SAI was done with a FORTRAN program. A Macintosh Filemaker
Pro template is now available for scoring.
Munbys (1983) report on the use of the SAI was a criticism of its use and of the validity
of the instrument itself. Munby also reported conflicting reports from various researchers using
the SAI. Baker (1985) stated that Munbys criticism leaves the researcher in a quandary, noting that researchers can spend a great deal of time developing their own instrument to assess
attitudes toward science, avoiding all of the conceptual problems inherent in previous instruments, or the variable of attitude can cease to be a part of any study. Neither of these approaches
seems to be realistic. At best we can hope that a better instrument will be developed and that
all conclusions about attitude toward science can be regarded as tentative (p. 105).
Nagy (1978) investigated the SAI to look for empirical support for the distinction between
feelings and beliefs in a scientific attitude scale (p. 355). Assumptions underlying this investigation were stated: Since beliefs require cognitive learning, it is reasonable to expect that
feelings will correlate with achievement more strongly than beliefs unless beliefs are part of the
329
course content (p. 355). Nagy found that there were problems with some of the more difficult
vocabulary required by the SAI: for example, the words phenomena, objective, and idea-generating seemed to be cause for concern. In addition, Nagy built a case for using a five-choice
response format with an undecided or neutral category as opposed to the original four-choice
format without a neutral category.
It is important to pay attention to serious criticisms and suggestions regarding vocabulary,
item difficulty, and format in a revision of an instrument. This revision is an effort to do that.
Need for Revision
Many events have transpired since the SAI was developed. Many NSF-funded curricula and
various projects for teachers have had various impacts upon the teaching of science. Some of
these projects are in a third or fourth revision. Some of the science education work of the 1960s
and 1970s has been incorporated into text series produced by different publishers. Some of the
investigators of that period are still at work on extensions of those efforts. In addition, the AAAS
(1990) embarked on a project to develop a scientifically literate society by the year 2061.
Changes in direction and focus usually accompany changes in projects. This is a good reason
to update an instrument that may be used in support of both continuing efforts and new initiatives.
A reading of the items in the SAI provides an obvious reason for revision. The words he,
him, and his appear nine times, the words she, her, and hers are not used at all. Also, there are
two references to man. This use of language reflects the thinking of the era in which the instrument was developed. It also is cause for outright rejection of the use of the instrument today. This is sufficient reason to revise the SAI. Also, as indicated above, Nagy (1978) discussed
some of the vocabulary used in the original instrument, and suggested revisions which have been
addressed (e.g., as noted above, items containing terms such as idea-generating activity, objective, and phenomena, when used with a five-choice format, produced an inordinate number of undecided choices when compared with the remainder of the SAI).
Method
The SAI revision is referred to as the SAI II. Since there is evidence for the content validity of the items in the original instrument with respect to the 12 position statements, we decided to make as few changes as possible while responding to criticisms and suggestions. This
evidence was presented in item selection and the field test of the original SAI. The primary goal
for the revision was to prepare an instrument which would be as widely useful as possible
and still retain the validity of the original SAI. To this end, we decided to focus on three goals:
(a) to eliminate gender-biased references, (b) to eliminate words that have been criticized as difficult for readers to understand, and (c) to shorten the instrument to make it easier to use.
All 12 of the original position statements were retained to represent the universe of content, attitudes toward science. In addition, there was no attempt to generate new statements. The
60 attitude statementsfive items for each of 12 position statements, 6 opposing positive and
negative statementswere revised for the purpose of making them easier to read. Difficult
words were eliminated, and the attitude statements were shortened when possible. After the attitude statements were edited, two attitude statements were eliminated for each scale except for
the last pair of positive and negative scales, which have to do with the individuals desire to do
scientific work. Thus, the SAI II has 40 Likert-type attitude statements instead of the 60 items
in the original SAI. In addition, we decided to use the five-point scale in a field test of the SAI
330
II on the basis of suggestions by individuals who have used the original SAI. The position statements and the corresponding attitude statements for the SAI II are presented in the Appendix.
The numbers by each item in the Appendix indicate the order of the items in the instrument.
Of the 60 attitude statements of the SAI, 40 were selected for use in the SAI II. Ten of the
40 statements were not changed. The 30 that were changed were changed to eliminate gender
bias or make the item easier to read. The following is an example of a change to eliminate gender-biased language (the revision is read without the portion lined through):
If one scientist says a theory is true, all other scientists will believe him it.
Some changes met both objectives of eliminating gender bias and making the item easier to
read:
If a scientist cannot answer a question, all he has to do is to ask another scientist can.
Scientific questions are answered by When one asks questions in science, he gets information by observing natural phenomena things.
One of the most important jobs of a Scientists must is to report exactly what they observe.
his senses tell him.
There are 12 position statements. Six positions are positive and are labeled 1-A through 6A. Six are negative and are labeled 1-B through 6-B. The A and B pair for each position are opposites of each other. The useful scales for analysis are 1-AB through 6-AB for each position
and the positive and negative scales consisting of 1-A through 6-A and 1-B through 6-B, respectively.
Scoring
The SAI II is scored by assigning point values to each of the attitude items. Point values
are assigned as shown in Table 1.
Scores for the various subscales can be determined by adding the scores for the respective
items. Scores may be determined for the 12 subscales, a total for the positive items, a total for
the negative items, and a total for the entire SAI II. The range of scores for each of the scales
1-A through 5-B is 315 (15 points 3 3 items). The range of scores for scales 6A and 6B is
Table 1
Point values for positive items and for negative items
Strongly agree
Mildly agree
12 neutral/undecided
Mildly disagree
Strongly disagree
Positive Items
Negative Items
5
4
3
2
1
1
2
3
4
5
331
525 (15 points 3 5 items). The range of scores for the entire SAI is 40200 (15 points 3
40 items). Refer to the Appendix for the various scales. The authors handled the data for the
field test entirely by machine after identification numbers were coded on students scanner
forms. An optical scanner read the data into a mainframe computer. The data were downloaded
from the mainframe to a personal computer. The downloaded data had to be cleaned up to eliminate mainframe reference numbers that were of no importance to the study and to parse the data
so they could be used by the scoring program. Microsoft Excel was used to remove the unwanted mainframe reference numbers and parse the data. Then the data was exported from Excel and imported into Filemaker Pro. A Filemaker Pro template was used to score each students
responses and to generate scores for each of the scales for each student. The Filemaker Pro template is available from the authors.
Field Test of the SAI II
Responses were collected from 588 students in 6th, 9th, and 12th grades in a rural/suburban middle school and high school in the same school district to determine how students at the
various grade levels would respond to the revised SAI. A total of 557 of the respondents provided usable papers. The responses of the top and bottom 27% of respondents on the total instrument have been compared on the various subscales to provide evidence for the validity of
the revised SAI. The rationale is that if there is a difference between the scores of the top scorers and the bottom scorers in favor of the top scores on the various subscales, those scales contribute to the instruments ability to distinguish between those with strong attitudes toward science and those with weak attitudes toward science. This method of analyzing Likert data has
been used by others for science attitude scale development (Hassan, 1984; Koballa, 1984; Misiti, 1991). This work indicates that each of the various subscales contributes positively to the
total score for the instrument.
The SAI II has a score range of 40200. The number of items in each scale, the scale mean,
the scale standard deviation, the low and high score for this sample, and the score range for each
of the scales in presented in Table 2.
When the top and bottom 27% of scorers on the total test are compared on each of the six
subscales, scales 1-AB through 6-AB, a significant t was obtained in each case at the .05 level
of significance. The results of these t tests for independent samples are presented in Table 3.
Table 2
Mean (M) standard deviation (SD), low and high score, and range for each scale
of the SAI II for 557 respondents
Scale
1-AB
2-AB
3-AB
4-AB
5-AB
6-AB
Positive
Negative
Total
Score
No. of
Items
SD
Low
High
Scale Range
Range
6
6
6
6
6
10
20
20
40
22.0
23.4
23.0
18.3
21.6
31.0
73.5
67.6
141.2
2.95
3.48
3.15
2.56
3.82
8.74
10.00
8.66
15.07
9
12
12
10
6
10
40
36
97
30
30
30
28
30
50
96
93
182
630
630
630
630
630
1050
20100
20100
40 200
332
Table 3
Comparison of total score high and low scorers on subscales 1-AB through 6-AB,
and positive scales, negative scales, and total score
High
Scale
1-AB
2-AB
3-AB
4-AB
5-AB
6-AB
Positive
Negative
Total
Low
SD
SD
p , .05
24.0
25.3
25.9
18.9
24.4
41.0
83.2
76.2
159.4
2.42
2.79
2.40
2.67
3.06
5.74
5.71
5.70
6.93
20.0
21.4
21.6
17.8
18.3
23.2
62.8
59.4
122.3
2.58
3.58
3.34
2.65
3.42
5.99
7.63
7.18
7.33
13.81
10.50
12.72
3.47
16.28
26.29
26.23
22.27
45.03
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
These data demonstrate that the instrument distinguishes between those who score high and
those who score low on the total score, and are useful in establishing the validity of the SAI II.
Validity
The statements in the SAI II are essentially the same as those in the SAI. They have been
revised to improve readability and eliminate gender-biased language as indicated above. Therefore, content validity is claimed on the basis of the original judgments of the panel of judges regarding the relation of the attitude items to the attitude positions. Construct validity of the SAI
was demonstrated in the original field test.
An attempt was made to obtain support for the validity of the SAI II with confirmatory factor analyses of the data from the 557 respondents. One approach to the analysis involved all
40 items. In another analysis, the group of intellectual scales (1-A through 4-B) were examined
separately from the emotional scales (4-A through 6-B). The resulting factor loadings were not satisfactory in any of these efforts. Exploratory factor analysis was attempted to determine a factor
structure that might be supported by subsequent confirmatory factor analysis. This effort
produced five scales using 30 of the 40 items in the SAI II. However, the grouping of items
was not very satisfactory in terms of attempts to give meaning to the groupings. Establishing meaning for the groupings of items identified by factor loadings is an important step in establishing validity with factor analysis (Gable, 1993; Hatcher, 1994). In addition, regrouping the items would
virtually eliminate the support gained by judges for the validity of the instrument. Therefore, the
40-item SAI II is being advanced as presented here without the support of factor analysis.
We have presented the results of an administration of the SAI II to 588 students (557 useful respondents), which indicates that the scales of the instrument distinguish between those who
have more positive attitudes toward science and those who have less positive attitudes toward
science as determined by the total score on the SAI II. The t-test comparisons of the high and
low scorers is evidence that the various subscales contribute positively to the total score of the
instrument. This is evidence that the various scales are related to each other and measure something similar. Coupled with judgments that the items of the instrument are related to the scientific attitudes it is supposed to assess, validity is claimed for the SAI II.
Considering the wide use of the SAI, it is possible that construct validity for the SAI II will
be demonstrated further as it is used in a variety of studies. There have been many requests in
333
recent years by individuals desiring to use the SAI and to modify it in various ways. There clearly is a need for a revision of this instrument to be available for continuing research regarding
scientific attitudes.
Reliability
A split-half reliability coefficient was computed for the entire group of 557 respondents.
Application of the Spearman Brown correction for split-half reliability to the correlation coefficient yields a reliability coefficient of .805. Cronbachs alpha reliability coefficient is .781 for
this group.
Discussion
Validity support from factor analysis would have helped establish the viability of this instrument and would have been a helpful response to previous criticism. On the other hand, Lederman (1992) examined various efforts to assess both teachers and students attitudes toward
and views about science. He noted that science teachers do not possess adequate conceptions
of the nature of science, irrespective of the instrument used to assess understandings (p. 345).
If science teachers are lacking in this area, one might expect that while scales designed to assess attitudes can distinguish among those who score high and those who score low on the total instrument, it is possible that the objects of the scales are so ill-formed in the subjects, students in this case, that we are not able to use their responses to confirm the scales as factors.
We believe this version of the SAI, the SAI II, is a significant improvement over the original. It is hoped that this update of a widely used instrument will provide additional interest in
this important dimension of science, and will promote advances in assessing science attitudes
and distinguishing among attitudes, values, and beliefs.
Permission to use the SAI II for research purposes may be obtained by writing to the authors. A copy
of the instrument will be provided to any interested person. Please address correspondence to: Richard W.
Moore, 301 McGuffey Hall, Miami University, Oxford, OH 45056; or e-mail [email protected]. A
Macintosh Filemaker Pro template for scoring also is available from this author at no charge. A revision
of the Science Teaching Attitude Scales (STAS) (Moore, 1973), which incorporates revisions in the SAI,
also is available. Studies involving attitudes of both teachers and their students may find these instruments
to be useful.
Appendix
Position Statements and Attitude Statements of the SAI II
These are the position statements and corresponding attitude statements of the SAI II.
The position statements are labeled with a number and a letter: for example, 1-A. The letter designates whether the position statement is positive (A) or negative (B). The position statements are in pairs, where the pair 1-A and 1-B are intended to be opposite positions regarding
the same point of view. The numbers in front of each attitude statement indicates its number in
the SAI II.
1-A. The laws and/or theories of science are approximations of truth and are subject to
change.
4. Scientists are always interested in better explanations of things.
334
335
6-A. Being a scientist or working in a job requiring scientific knowledge and thinking
would be a very interesting and rewarding lifes work. I would like to do scientific
work.
1. I would enjoy studying science.
27. I would like to work with other scientists to solve scientific problems.
30. I may not make great discoveries, but working in science would be fun.
36. I would like to be a scientist.
40. Working in a science laboratory would be fun.
6-B. Being a scientist or working in a job requiring scientific knowledge and thinking
would be dull and uninteresting; it is only for highly intelligent people who are willing to spend most of their time at work. I would not like to do scientific work.
13. The search for scientific knowledge would be boring.
14. Scientific work would be too hard for me.
22. I do not want to be a scientist.
37. Scientists do not have enough time for their families or for fun.
39. Scientists have to study too much.
References
American Association for the Advancement of Science. (1990). Science for all Americans.
Washington, DC: Author.
Baker, D.R. (1985). Predictive value of attitude, cognitive ability, and personality to science achievement in the middle school. Journal of Research in Science Teaching, 22, 103113.
Finson, K.D., & Enochs, L.G. (1987). Student attitudes toward science-technology-society
resulting from visitation to a science-technology museum. Journal of Research in Science
Teaching, 24, 593609.
Gable, R.K., & Wolf, M.B. (1993). Instrument development in the affective domain (2nd
ed.). Boston: Kluwer Academic.
Hassan, A.M.A., & Shrigley, R.L. (1984). Designing a Likert scale to measure chemistry
attitudes. School Science and Mathematics, 84, 659669.
Hatcher, L. (1994). A step-by-step approach to using the SAS system for factor analysis and
structural equation modeling. Cary, NC: SAS Institute.
Kerlinger, F.N. (1986). Foundations of behavioral research (3rd ed.). New York: Holt,
Rinehart, and Winston.
Koballa, T.R. (1984). Designing a Likert-type scale to assess attitude toward energy conservation: A nine step process. Journal of Research in Science Teaching, 20, 709723.
Koballa, T.R. (1988). Attitude and related concepts in science education. Science Education, 72, 115126.
LaShier, W.S., & Nieft, J.W. (1975). The effects of an individualized, self-paced science
program on selected teacher, classroom and student variablesISCS level one. Journal of Research in Science Teaching, 12, 359369.
Lederman, N.G. (1992). Students and teachers conceptions of the nature of science: A review of the research. Journal of Research in Science Teaching, 29, 331359.
Mead, M., & Metreaux, R. (1957). Image of the scientist among high school students. Science, 126, 384390.
Misiti, F.L., Shrigley, R.L., & Hanson, L. (1991). Science attitude scale for middle school
students. Science Education, 75, 525540.
Moore, R.W. (1973). The development, field test, and validation of scales to assess teachers attitudes toward teaching elementary school science. Science Education, 57, 271278.
336
Moore, R.W., & Sutman, F.X. (1970). The development, field test, and validation of an inventory of scientific attitudes. Journal of Research in Science Teaching, 7, 8594.
Munby, H. (1983). Thirty studies involving the Scientific Attitude Inventory: What confidence can we have in this instrument? Journal of Research in Science Teaching, 20, 141162.
Nagy, P. (1978). Subtest formation by cluster analysis of the Scientific Attitude Inventory,
Journal of Research in Science Teaching, 15, 355360.
Noll, V.H. (1935). Measuring scientific attitude. Journal of Abnormal and Social Psychology, 30, 145154.
Shrigley, R.L., Koballa, T.R., & Simpson, R.D. (1988). Defining attitude for science educators. Journal of Research in Science Teaching, 25, 659678.
Welch, W.W. (1972). Evaluation of the PSNS course, II: Results. Journal of Research in
Science Teaching, 9, 147156.
Willson, V.L., & Lawrenz, F. (1980). Relationship between teacher preparation in NSF institutes and student attitudes and perception of the classroom learning environment. Journal of
Research in Science Teaching, 17, 289294.