1
Henry Jay Becker, Internet Use by Teachers: Conditions of Professional Use and Teacher-Directed Student Use. Report #1, Teaching, Learning, and Computing: 1998 National Survey. University of California, Irvine. February, 1999.
http://www.crito.uci.edu/TLC/FINDINGS/internet-use/startpage.htm
3
In all tables, (N) refers to the actual number of teachers responding in that category. Percentages and means however, reflect weights based on the inverse of the probability that each teacher has selected for the sample. Two different weights were used, depending on whether the table refers to data from only the national probability sample of schools (as in Table 1) or from teachers in both the national probability sample and the two categories of purposive samples of schools ("educational reform" and "high-end technology") that comprised 45% of the total set of schools studied. For further information about the sample and data collection methodology, see
Appendix B at the end of this document.
4 In Table 2, we focus on computer use in one class taught by each sampled teacher, which we call the 'selected' class. This is the class where the teacher felt most accomplished in teaching. In this table, the teachers themselves are defined not in terms of the subject that they taught most often, but by which subject they taught that particular class.
5 In Table 4, subject categories were collapsed and teachers from the purposive samples were included in order that each cell in the table was based on at least 30 (actually 29) cases.
6 The difference in the percentage of frequent users among computer-assigning teachers is only 12 percentage points between elementary teachers with at least one computer in their classroom, but fewer than one-per-four students (55%); and elementary teachers with a better computer-student ratio (67%).
7 Note that "graphing software" was not a category used in the survey. However, graphing programs were mentioned by relatively few math teachers in an open-ended question about the specific software they found to be most valuable with their students.
8 The question about "best programs used by students" was asked of 50% of the survey sample. The question about the "most valuable" software was asked of the other 50%. The latter question, though, was asked differently—it incorporated both teacher use and student use and inquired separately about each of the past five years. The analysis in this section incorporates answers from both questions, except that for the second group, only reports about the past two years are included and only when the software appeared to be used by students rather than the teacher herself.
9 K-Means clustering (or "Quick Cluster" in SPSS).
10 In order to make use of all teachers' data, the percentages and rates in this
section come from the full set of TLC schools—that is, both the probability sample and the various purposively selected reform-involved and high-technology schools. However, in many respects, differences among these two
samples of teachers are hardly noticeable. For example, the percentage of teachers who are in clusters 2 through 10 combined (whom we might call "regular computer-assigning teachers") is nearly the same in the probability
sample as it is among the teachers in the reform-plus-high-tech TLC schools (the purposive sample). For example, 54% of probability-sample elementary teachers regularly assign computer work compared to 56% of teachers in the
elementary schools in the purposive sample. (Comparable percentages for middle school teachers are 37% for each sample; and for high school teachers, 42% in the probability sample and 39% in the purposive sample.)
11 An important reservation to this claim should be noted: A large fraction of students' more intensive computer experiences occur in computer education and business education classes, and teachers of those classes are less likely than average to prioritize improvement of student writing as a primary objective. (See text below)
12 We recognize that teachers might have different objectives for different types of software that they use. Unfortunately our data about teachers' objectives was not specific to one or another type of software but just to how a particular teacher used computers with her students. The consequence of this data limitation is that if we find differences in objectives by software type, it is likely that the true differences are even larger than shown, since the teacher's objectives at best are "averaged" across different software that she uses.
13 "Use" was defined as students using that type of software in at least three lessons during the year. Rather than employing the absolute difference in percent using between objective-holders and non-holders, that difference was divided by the combined sample standard deviation of "use/non-use," in effect creating an "effect size" measurement. This was to take into account the fact that differences between two very small (or very large) percentages need not be as large as differences between two modestly sized percentages for an equivalent "effect" to be registered. Effect sizes greater than .20 are show in the table.
14 The pattern in Table 13 is somewhat different than the pattern in Table 12, and several reasons are likely contributors. First, Table 13 uses a criterion of "frequent" use of a given type of software while Table 12 is based on the percentages who use that software even occasionally. Second, Table 13 uses the purposive (reform plus high tech) samples as well as the probability sample, while Table 12 only uses the latter. Third, different statistics are used in each, for different purposes. In Table 12, the focus is on contrasting teachers who prioritize certain objectives with those who prioritize others, and leads to an effect size calculation. In Table 13, the focus is on comparing frequent users of different types of software (ignoring those who are not frequent users of any type of software) and attends to differences in the kinds of objectives these contrasting groups of frequent software-using teachers have.
15 This is based on a scale which counted weekly or greater use of computers for keeping grades, making handouts, writing lesson plans, and getting information from the Web, and occasional (or greater) use for parent correspondence, use of camcorder/scanner/cameras, exchanges of files with other teachers, and World Wide Web posting.
16 There is a three-fold difference in r-squared, .09 vs. .27.
17 Ronald E. Anderson and Amy Ronnkvist, "The Presence of Computers in American Schools," Report #2, Teaching, Learning, and Computing: 1998 National Survey. University of California, Irvine. June, 1998. http://www.crito.uci.edu/tlc/findings/computers_in_american_schools/html/startpa ge.htm