FINDINGS Responding to a Global Pandemic: The Role of K-12 Science Teachers (COVID 2020)
Findings
Sources of Information About COVID
To find out where teachers accessed information about COVID, the survey asked them what media sources they consulted and how useful they found those sources. Findings regarding sources of information about COVID, combined before and after schools closed, are described below.
Whether or not they taught about COVID, more than two-thirds of elementary teachers relied on the CDC website and local news stations to a substantial extent (i.e., those responding 3 or 4 on a four-point scale ranging from 1 “Not at all” to 4 “To a great extent”) as sources of information about COVID. However, the CDC website was more likely to be accessed by elementary teachers who taught about COVID than those who did not (84 vs. 73 percent). Elementary teachers who taught about COVID were also more likely than those who did not to access other health organization websites to a substantial extent, including the National Institutes of Health (NIH) website, (42 vs. 27 percent) and the Johns Hopkins Coronavirus Resource Center website (41 vs. 25 percent).
Large percentages of teachers at each grade band relied on health information websites to learn about COVID. Teachers who taught about COVID were generally more likely to access health information websites to a substantial extent than teachers who did not teach about COVID.
At the middle and high school levels, teachers also accessed a variety of media sources to find information about COVID (see Tables 3 and 4). However, there were many differences between teachers who taught about COVID and those who did not in the extent to which these media sources were utilized. For example, middle school teachers who taught about COVID were more likely than those who did not to access the CDC website (91 vs. 79 percent), WHO website (62 vs. 40 percent), NIH website (58 vs. 41 percent), popular science magazines (55 vs. 28 percent), and the Johns Hopkins Coronavirus Resource Center website (54 vs. 39 percent) to a substantial extent. Similarly, high school teachers who taught about COVID were more likely than those who did not to access the CDC website (90 vs. 80 percent), Johns Hopkins Coronavirus Resource Center website (67 vs. 41 percent), NIH website (66 vs. 46 percent), WHO website (66 vs. 40 percent), newspapers (60 vs. 45 percent), and popular science magazines (60 vs. 45 percent) to a substantial extent.
Teachers were increasingly likely to rely on written news sources and health/science organization websites with increasing grade range. These sources were also more likely to be accessed by teachers who taught about COVID than those who did not.
The items shown in Tables 2–4 were combined into five composite variables: (1) local/national television news stations, (2) online news/social media, (3) written news sources, (4) health/science organization websites, and (5) personal conversations. As can be seen in Table 5, as grade range increased, teachers were more likely to rely on written news sources (34, 43, and 49, respectively) and health/science organization websites (46, 59, and 64, respectively). Additionally, there were significant differences between teachers who taught about COVID and those who did not. Written news sources were more likely to be accessed by middle and high school teachers who taught about COVID than those who did not. Similarly, elementary, middle, and high school teachers who taught about COVID were more likely to utilize health/science organization websites than those who did not teach about COVID.
There were some differences in the types of media teachers consulted for information about COVID by the political leaning of the county.
The five composite variables related to teachers’ sources of information about COVID (shown in Table 5) were examined by equity factors. As can be seen in Table 6, two differences were noted based on political leaning, although the magnitude of these differences is small. Teachers in Democratic-leaning counties were more likely than those in Republican-leaning counties to rely on online news/social media (32 vs. 28) and written news sources (45 vs. 38). Looking at these data by community type, teachers in rural schools were more likely than those in suburban or urban schools to rely on written news sources (45, 42, and 38, respectively). Additionally, teachers in schools in the highest URM quartile were more likely than those in the lowest URM quartile to rely on local or national television news stations as a source of information about COVID (54 vs. 48).
Teaching About COVID
The survey asked teachers whether they devoted class time to COVID, both before and after school buildings closed, and what influenced their decision. Teachers who addressed COVID were also asked how much class time they devoted, what instructional activities were used, and what specific topics of the virus/disease were addressed. These data are discussed in this section of the report.
Large proportions of teachers at each grade band and across all equity factors devoted class time to COVID.
As can be seen in Table 7, three-fourths or more of teachers at each grade band devoted class time to COVID. Life science teachers were more likely than non-life science teachers to address COVID at both the middle and high school levels (84 vs. 73 percent and 92 vs. 69 percent, respectively). However, it is noteworthy that approximately 7 in 10 non-life science teachers at the middle and high school levels also took up the topic.
Table 8 shows percentages of classes in which COVID was addressed by various equity factors. Large percentages of teachers addressed COVID across equity factors, with no significant differences observed.
Of those teachers who devoted class time to COVID, most spent more than one class session. Life science teachers at the high school level were more likely than non-life science teachers to devote more than three class sessions to the topic.
At each grade band, a majority of teachers who taught about COVID devoted more than one class session to the topic (see Table 9). Additionally, the distribution of class sessions is significantly different between life science and non-life science classes at the high school level. This difference can potentially be attributed to the fact that life science teachers were more likely than non-life science teachers to spend three or more class sessions on COVID.
The numbers of class sessions devoted to COVID were generally equitably distributed.
Analyses were conducted to examine COVID-focused class time by equity factors. As can be seen in Table 10, few differences were apparent, suggesting that student access to instruction focused on COVID was generally equitably distributed. However, there was one notable difference. The distribution of class sessions is significantly different between the highest and lowest FRL quartiles, likely because classes in high-poverty schools were more likely than those in low-poverty schools to devote more than three class sessions to COVID.
Large percentages of life science teachers addressed COVID as part of their curriculum. Elementary teachers and non-life science teachers were more likely to address COVID as a standalone topic.
When they addressed COVID, over 80 percent of life science teachers at the middle and high school levels reported doing so as part of their curriculum (see Table 11). Conversely, non-life science teachers were more likely than their life science counterparts to treat COVID as a standalone topic, unrelated to the rest of their science curriculum. At the elementary level, teachers were more likely to address COVID as a standalone topic than as part of their curriculum.
Regardless of grade range, over 80 percent of teachers indicated that their students asked about COVID before they addressed it.
The survey included an item regarding whether students asked about COVID before teachers began addressing it. Across grade ranges, nearly 80 percent of teachers who devoted some class time to COVID indicated that their students asked questions about it first (see Table 12). There was no significant difference between life science classes and non-life science classes at the middle or high school level, suggesting that students turned to science teachers in all disciplines for information about COVID.
Differences in percentages of students who asked about COVID before teachers addressed it were seen based on political leaning of the county and URM quartile of the school.
Although large percentages of teachers who devoted some class time to COVID indicated that their students asked questions about it first, there were a few differences by equity factors. As can be seen in Table 13, teachers in schools in the highest URM quartile were more likely than those in the lowest URM quartile to indicate that students asked questions about COVID before they addressed it (85 vs. 76 percent). In addition, teachers in Democratic-leaning counties were more likely to report that students asked questions about COVID before they addressed it than teachers in Republican-leaning counties (81 vs. 76 percent).
There was a great deal of variation in the types of questions students asked about COVID.
As previously mentioned, student questions played a large part in teachers’ response to COVID. To learn more about these questions, the survey asked teachers to list the five most common questions they received, which researchers coded into categories. As can be seen in Table 14, there was a great deal of variation in the types of questions students asked at all three grade bands, including preventing transmission, changes to schools, risk of infection, and what the virus/disease is. However, it appears that students generally asked similar types of questions in life science and non-life science classes.
Large majorities of teachers would have addressed COVID even if their students had not asked, regardless of grade range, teaching assignment (life/non-life) and equity factors.
As can be seen in Table 15, over 80 percent of teachers at each grade range said they would have addressed COVID even if students had not asked about it. At the middle and high school levels, life science and non-life science teachers were equally likely to indicate that they would have addressed COVID even if students had not asked. This is a particularly striking finding given that viruses are usually not a topic of instruction in non-life science classes.
Teachers’ inclination to address COVID were also examined by equity factors, with only one significant difference apparent. As can be seen in Table 16, teachers in Democratic-leaning counties were more likely than teachers in Republican-leaning counties to indicate they would have addressed COVID even if students had not asked (90 vs. 85 percent).
While school buildings were open, teachers utilized class discussions driven by student questions to teach about COVID. After school buildings closed, teachers increasingly relied on readings and videos.
The survey presented teachers with a list of instructional activities that could have been used to address COVID. As can be seen in Tables 17–19, regardless of grade range or focus of the class (life science or not), the most prevalent instructional activities used by teachers while school buildings were open were answering student questions (ranging from 85 to 87 percent of teachers) and whole class discussion (ranging from 72 to 74 percent of teachers).
Interestingly, after school buildings closed, instructional activities used to address COVID were quite different from before school buildings closed. Although teachers across grade levels still devoted class time to answering student questions, this activity was less prevalent than before school buildings closed (87 vs. 73 percent at the elementary level, 87 vs. 60 percent at the middle school level, and 87 vs. 59 percent at the high school level). Whole class discussions and teacher lectures were also less likely to occur at each grade band after school buildings closed, likely due to the shift to online instruction and limited/disrupted instructional time.
Conversely, teachers at each grade band were more likely to have students read about COVID after school buildings closed. At the high school level, teachers were also more likely to have students watch videos about COVID after school buildings closed, an activity that was more common in life than non-life science classes. This shift toward readings and videos is likely due to the fact that these activities are amenable to an online format. Further, it is possible that readings and videos about COVID became more prevalent and readily available to teachers over time.
Composite mean scores indicate that teachers were generally unlikely to use active learning strategies to teach about COVID, relying more on group/whole class discussions.
The items in Tables 17–19 were combined into three composite variables: (1) Group/Whole Class Discussions, (2) Individual Active Learning Strategies, and (3) Individual Passive Learning Strategies. As can be seen in Table 20, composite means suggest that teachers across grade bands were moderately likely to use group/whole class discussions to address COVID but unlikely to use active learning strategies (e.g., do a hands-on/laboratory investigation about COVID). The composite means also reveal differences in teachers’ use of instructional activities at the secondary level. Middle and high school life science teachers were more likely than their non-life science counterparts to use passive learning strategies (40 vs. 33 and 46 vs. 28, respectively). High school life science teachers were also more likely than non-life science teachers to address COVID via active learning strategies (22 vs. 14).
Teachers of classes in high-poverty schools were more likely to use group/whole class discussions and passive learning strategies than teachers in low-poverty schools.
Looking at these composites by equity factors reveals some significant differences by FRL. As can be seen in Table 21, teachers of classes in high-poverty schools were more likely than teachers of classes in low-poverty schools to utilize group/whole class discussions (65 vs. 59) and passive learning strategies (39 vs. 30). Although the use of active learning strategies was uncommon across quartiles, it was more likely to occur in classes in high-poverty schools than classes in low-poverty schools (18 vs. 13).
Across grade bands, the most commonly addressed topics while school buildings were open included ways to prevent coronavirus transmission, how coronavirus is transmitted, and what coronavirus/COVID-19 is. After school buildings closed, there was an increased focus on topics related to public health, including the impacts of social distancing and factors that place people at risk.
The survey provided a list of topics teachers could have addressed during COVID instruction while their school building was still open (see Tables 22–24). Across grade ranges, the topics most commonly addressed were ways to prevent coronavirus transmission from one individual to another, how coronavirus is transmitted among humans, and what coronavirus/COVID-19 is. Additional topics were also frequently addressed at the high school level, including symptoms of COVID-19 (66 percent), where coronavirus originated (64 percent), likelihood that coronavirus/COVID-19 would spread throughout the United States (56 percent), and factors that place people at risk for contracting coronavirus (54 percent).
After school buildings closed, many of the same topics continued to be addressed in instruction across grade bands. However, topics with a public health focus became increasingly prominent, including the impacts of social distancing and factors that place people at risk for contracting coronavirus.
Several differences in topics addressed were apparent when comparing life and non-life science teachers at the secondary level. Life science teachers were more likely than their non-life science counterparts to address a range of topics, both before and school buildings closed. These differences were also generally larger at the high school level than at the middle school level.
Elementary teachers were less likely than middle or high school teachers to address COVID transmission, COVID treatment/diagnosis, or further/advanced COVID topics. Further, high school life science teachers were more likely than non-life science teachers to cover each topic area.
The items shown in Tables 22–24 were combined into three composite variables: (1) COVID transmission, (2) COVID treatment/diagnosis, and (3) Further/advanced COVID topics. Composite means shown in Table 25 indicate that COVID transmission was commonly addressed in instruction across grade bands (composite mean scores ranging from 79 to 87). However, these data also point to differences in the focus and scope of instruction by grade band and teaching assignment. Elementary teachers were less likely than middle and high school teachers to address COVID transmission (79, 84, and 87, respectively), COVID treatment/diagnosis (13, 25, and 36, respectively) and further/advanced COVID topics (31, 50, and 62, respectively). Similarly, at the high school level, teachers of life science classes were more likely than teachers of non-life science classes to cover COVID transmission (92 vs. 80), COVID treatment/diagnosis (45 vs. 23) and further/advanced COVID topics (65 vs. 57).
There were significant differences in topics addressed during COVID instruction based on FRL quartile, each of which favored classes in high-poverty schools.
The composites related to topics addressed during COVID instruction were also examined by equity factors. As can be seen in Table 26, there were significant differences based on FRL quartile, each in favor of classes in high-poverty schools. Teachers in high-poverty schools were more likely than teachers in low-poverty schools to address COVID transmission (87 vs. 82), COVID treatment/diagnosis (31 vs. 22), and further/advanced COVID topics (56 vs. 46).
Elementary teachers tended to use commercially published materials to teach about COVID, while middle and high school teachers relied heavily on units and lessons they created.
Teachers were asked to think about the instructional materials they utilized to teach about COVID. As can be seen in Table 27, 87 percent of elementary teachers relied on commercially published materials, and over half used units or lessons they created. At the secondary level, teachers were even more likely to create their own units or lessons for teaching about COVID, as 65 percent of middle school teachers and 70 percent of high school teachers reported using self-created instructional materials (see Tables 28 and 29).
Student Engagement
Teachers and students had to navigate a very atypical set of circumstances after school buildings closed, many of which altered the way teachers taught and the way students learned. Therefore, the survey asked teachers to reflect on student engagement during this time. Findings pertaining to teachers’ impressions of student engagement after school buildings closed are discussed in this section of the report.
Student engagement across grade bands was substantially lower after school buildings closed.
As might be expected, over 80 percent of teachers indicated that student engagement was substantially lower after school buildings closed (see Table 30). This finding was consistent across grade bands and teaching assignments.
Differences in student engagement were apparent based on community type and political leaning of the county.
Teachers’ impressions of student engagement after school buildings closed were also examined by equity factors. As can be seen in Table 31, teachers in schools in urban settings were more likely to rate student engagement as substantially lower after school buildings closed than teachers in suburban or rural settings (82, 87, and 90 percent, respectively). Further teachers in Republican-leaning counties were more likely than teachers in Democratic-leaning counties to report substantially lower student engagement after school buildings closed (90 vs. 84 percent).
Teacher Decision Making
A major focus of the survey was identifying factors that influenced teachers’ decisions regarding whether to address COVID or not. Teachers were presented with open-ended items that asked them to identify the most important reasons why they either did or did not address COVID in their instruction. Additionally, teachers were asked to respond to a set of survey items aligned with the TPB. Data about teacher decision making are discussed in this section of the report.
The most common reason why middle and high school teachers taught about COVID was that it was a relevant and current event. The most common reason why elementary teachers taught about COVID was to address student fear and anxiety.
Teachers who addressed COVID were asked to list the most important reasons why they decided to do so. Across grades bands, about 30 percent of teachers decided to teach about COVID because they felt that it was a relevant and current event for students (see Table 32). At the elementary level, about one-third of teachers taught about COVID in efforts to address student fears/anxiety, a reason that was less commonly given by middle or high school teachers. High school teachers were more likely than middle school or elementary teachers to address COVID because it was related to their standards or curriculum (20, 12, and 5 percent, respectively).
There were few differences in the most common reasons why teachers decided to address COVID by equity factors.
Examining these data by equity factors reveals few differences in the reasons why teachers addressed COVID (see Table 33). Teachers in schools in the highest URM quartile were less likely than teachers in schools in the lowest URM quartile to address COVID because it was a relevant/current event (25 vs. 34 percent). In addition, teachers in high-FRL schools were less likely than those in low-FRL schools to address COVID because it was related to their standards/curriculum course (8 vs. 16 percent).
The most common reason why middle and high school teachers did not address COVID was that the topic was not related to their standards or curriculum. Several reasons were common among elementary teachers, including the age of their students, lack of stable knowledge about COVID, and the desire to avoid causing fear/anxiety.
Teachers who did not teach about COVID were asked to list the most important reasons behind their decision. Interestingly, there was a great deal of overlap in the reasons why teachers did and did not teach about COVID. As can be seen in Table 34, the single most prominent reason why middle and high school teachers decided not to address COVID was because the topic was not related to their standards or curriculum (50 percent of high school teachers and 43 percent of middle school teachers). However, a mixture of factors appeared to steer elementary teachers away from teaching about COVID, including the age of their students (25 percent), lack of alignment to standards or curriculum (21 percent), lack of stable knowledge about COVID (16 percent), the desire to avoid causing fear/anxiety (15 percent), and that the topic was prohibited by their district or school (15 percent).
There was some variation in the most common reasons why teachers decided not to address COVID based on community type and political leaning of the county.
Examining the most common reasons why teachers did not teach about COVID by equity factors revealed some significant differences. Teachers in urban settings were more likely than teachers in suburban settings not to teach about COVID because it was not related to their standards or curriculum (55 vs. 44 percent). Additionally, teachers in rural settings were more likely than Horizon Research, Inc. 44 October 2021
those in urban settings to avoid teaching about COVID because they did not want to cause fear/anxiety for their students (20 vs. 6 percent). Differences were also seen based on political leaning. Teachers in Democratic-leaning counties were more likely than those in Republican-leaning counties to cite a lack of time to teach about COVID (19 vs. 9 percent). Conversely, teachers in Republican-leaning counties were more likely than teachers in Democratic-leaning counties to not teach about COVID because of a lack of stable knowledge about the virus/disease (23 vs. 9 percent).
Teacher attitudes toward teaching about COVID, perceptions of control over teaching about COVID, and feelings of self-efficacy had a substantial influence on whether they addressed COVID in their instruction.
As previously mentioned, the survey included several items aligned with the TPB. These items were intended to measure the extent to which various factors influenced whether teachers addressed COVID in their instruction. The items were combined into four composite variables:
- Attitude Towards Teaching About COVID6
- Teaching about COVID is undesirable/desirable.
- Teaching about COVID is the wrong thing to do/right thing to do
- Teaching about COVID is harmful/beneficial.
- Teaching about COVID is unimportant/important.
- Teaching about COVID is uninteresting/interesting.
- Teaching about COVID is a waste of my time/a good use of my time.
- Subjective Norm7
- I felt pressure from others to teach about coronavirus/COVID-19.
- It was expected that I teach about coronavirus/COVID-19.
- Self-Efficacy
- I am confident in my ability to successfully teach about coronavirus/COVID-19.
- It was difficult for me to teach about coronavirus/COVID-19.
- Control
- It was up to me whether or not to teach about coronavirus/COVID-19.
- The decision about whether or not to teach about coronavirus/COVID-19 was beyond my control.
The composite mean scores, shown in Table 36, highlight important aspects of teacher decision making. Across grade bands, attitude (composite means ranging from 74 to 83) and control beliefs (composite means ranging from 78 to 87) had a substantial influence on whether teachers taught about COVID, although both factors were more influential at the secondary level than at the elementary level. Self-efficacy also had a major influence on whether teachers taught about COVID at the middle and high school levels (composite means of 68 and 72, respectively), but less so at the elementary level (composite mean of 57). Conversely, subjective norms had very little influence on teacher decision making at any grade band.
6Items in this composite were ranked on a 7-point semantic differential scale, with the two polar opposite adjectives as the endpoints.
7Items in the following three composites were ranked on a 7-point scale from Strongly Disagree to Strongly Agree.
Using their reports of the number of days and class periods spent on COVID instruction before and after school buildings closed, teachers were grouped into five ordinal categories, with the lowest level being no COVID instruction and the highest being four or more days of COVID instruction. A path model was then constructed to test relationships among amount of instruction, TPB factors, and other variables of interest. Both direct and indirect relationships were tested. Figure 3 displays the path model with all variables, but only the statistically significant relationships are shown. Because the coefficients are not straightforward to interpret, line weights represent the relative strength of the relationships. Note that the strongest predictors are teachers’ attitudes toward COVID instruction and their sense of control. The other TPB factors (subjective norm and self-efficacy) are also predictors but not as strong. The subject of the class (life science vs. other) predicted amount of instruction both directly and indirectly through control and self-efficacy. Similarly, whether the teacher had taken life science classes beyond the introductory level in college predicted instruction directly and through self-efficacy. None of the equity factors predicted instruction directly, but both school locale and political leaning of the county had a weak, indirect influence through attitude.
There were some small differences by community type and political leaning in the extent to which various TPB factors influenced teachers’ COVID instruction, but they were small.
The TPB composites were also examined by equity factors (see Table 37). In no case did differences between groups amount to more than five percentage points, suggesting no substantial differences. However, some were statistically significant. For example, for teachers in high-URM schools, control tended to be less influential than for those in low-URM schools (mean scores of 80 and 85, respectively). In addition, for teachers in urban schools, attitude toward teaching about COVID appeared to be more influential than for teachers in rural schools (mean scores of 82 and 78, respectively). The same was true of self-efficacy (mean scores of 70 and 65 for urban and rural teachers, respectively).
