Abstract
In this study, we investigated how early childhood teachers’ perspectives on their enacted literacy instruction aligned with the perspectives of observers of that same instruction. Two master teachers and two researchers, all with early childhood expertise, observed and reported their perspectives of 45 instances of literacy instruction. These were examined for alignment across each other and with teachers’ descriptions of their thinking during the instruction. Participants’ perceptions of instruction tended to align, yet there were notable differences in perceptions about context and goals. Although we often found common ground among participants regarding the purpose of instruction, there were interesting variations across participants that highlighted the complexity of classroom processes, the value of teachers’ contextual knowledge and the multiple perspectives brought to bear on the same instance of instruction.
Standardized observation measures that account for the presence of specific classroom practices are an important tool in understanding instruction in early childhood education (Burchinal, 2018; Fuligni et al., 2012) and are key elements of reform efforts aimed at improving classroom instruction (Connors, 2016; Tout et al., 2010). For early childhood literacy instruction, the field depends upon standardized observation measures such as the Classroom Assessment Scoring System (CLASS; Pianta et al., 2008a), Early Language and Literacy Classroom Observation (ELLCO; Smith et al., 2002), the Individualizing Student Instruction Classroom Observation System (ISI; Pelatti et al., 2014) and the Teacher Behaviour Rating Scale (TBRS; Assel et al., 2008) to evaluate instruction and provide professional development (Al Otaiba et al., 2016; Cabell et al., 2013; Landry et al., 2011). However, teaching also includes a ‘hidden side of the work’ (Freeman, 2002), in which teachers engage in complex thinking, bringing together multiple sources of information that inform their intentions and decisions resulting in their visibly enacted practice (Cohen et al., 2003; Lampert, 2003; Shulman, 1987). Observations of instruction do not necessarily capture this non-visible component of instruction. Thus, what may be observed about practice externally may not reflect all that is occurring internally. The purpose of this study was to understand how teachers’ perspectives on their enacted literacy instruction aligned with the perspectives of external viewers of that same instruction, focusing on key stakeholders, teachers and researchers, in early childhood education.
What observation can capture
Standardized observation measures are designed to identify the presence of a variety of discrete practices deemed to be important for developing children’s literacy skills. The use of such observational measures has been critical for understanding, describing and improving practice in early childhood settings. For example, researchers have used measures to describe the frequency of specific literacy practices (e.g. Pelatti et al., 2014), the quality of the classroom environment (e.g. Dynia et al., 2016; Early et al., 2007) and the quality of language-related interaction (e.g. Justice et al., 2008; Pianta et al., 2008b). Many of these measures have also been used to examine teacher characteristics that might predict different practices (e.g. Early et al., 2007; Pianta et al., 2014; Schachter et al., 2016) as one means of identifying how to improve literacy-related instruction. Importantly, observational measures have been used as tools to support changes in early childhood teachers’ practice both by researchers (e.g. Cabell et al., 2013; Pianta et al., 2008) and practitioners (e.g. Grace et al., 2008; McNerney et al., 2006). Specifically, observation measures can inform professional development by identifying those practices on which teachers may need extra support.
Thus, the utility of such measures, or observation of practice in general, is high in the field of early childhood. In addition to being used for research and improving instruction, in the United States, such measures are also important components of Quality Improvement Systems (Connors and Morris, 2015) and are included in requirements for Head Start recertification (Improving Head Start for School Readiness Act, 2007). Observation is also critical for programme self-assessment and is a recommended practice for supporting teacher development (National Association for the Education of Young Children, 2011). In these cases, those observing instruction are often coaches, directors or co-teachers who might have various levels of training regarding conducting observation, including how to use and interpret the resulting data. Additionally, viewers may have differing purposes for using observation or observational tools, as well as differing background perspectives on teaching – all of which may shape how they conduct observations and what they glean from those observations.
What observation may not capture
As described above, observation and observational measures can identify discrete instructional practices as these are enacted by teachers. However, external observations of teachers’ instruction may not reflect what teachers are thinking while they enact instruction. Indeed, there is emerging evidence that teacher intentions and priorities may differ from those of external observers (Munby, 1982; Schachter, 2017). Furthermore, observations are not necessarily able to capture the intentionality informing teachers’ decisions to enact these practices. In other words, observation cannot explain how or why teachers choose to implement specific practices. Teachers bring multiple sources of information to bear as they make their instructional decisions, including information about the contexts in which they deliver instruction (Lampert, 2003; Lee, 2014; Shulman, 1987). The distinction between what is observed and what teachers are thinking may be important if standardized observational tools are relied upon to identify teaching practices and to evaluate and research early childhood settings. This could lead to problems in accurately identifying intended practices as these are enacted and could have important implications for the evaluation and research of early childhood settings in which standardized measures play a large role.
Indeed, the difference between intention and observation may in part contribute to the reported absence of literacy practices in early childhood classrooms (e.g. Justice et al., 2008; Pelatti et al., 2014). That is, we may not see teachers engaging in particular practices because viewers are not observing those practices in a way that is aligned with how teachers see themselves using those practices. Thus, it is important to understand how accurately external observations link to teachers’ intentions during enacted instruction.
Purpose
The purpose of this study was to understand how teachers’ perspectives on their enacted literacy instruction aligned with the perspectives of viewers of that same instruction. This is important for understanding how observation provides insights into classroom practice. To that end, we asked the following research question: How do early childhood teachers’ reported intentions during literacy instruction align with other early childhood teachers’ and researchers’ observations of that instruction?
Conceptual framework
In this study, we employed a phenomenological approach (Creswell, 2007; Marton, 1981), seeking to describe both an observable phenomenon as well as the participants’ experience of the phenomenon. This approach allowed us to focus on both what external viewers observed about early childhood teachers’ literacy instruction as well as the actual teachers’ thinking about that instruction (Schachter, 2017). We were able to understand both how the instruction was perceived externally as well as how it was perceived by those carrying out the instruction, thereby illuminating both observations and intentions.
Method
These data were collected as part of a broader study investigating teachers’ pedagogical reasoning and use of assessment data to inform instruction. For the purposes of this study, we focused on a subset of participants and data collected in the first data-collection phase. In the next sections, we will describe the participants and data-collection procedures for this specific study.
Participants
Study teachers
Twelve preschool teachers participating in the broader study also participated in the present study. The only requirement for participation was that teachers work in classrooms with children ages 3 to 5 years old and that they reported using informal or formal literacy assessments in their classrooms. These teachers, who we refer to as study teachers, were from various early childhood teaching contexts, including Head Start (n = 8), as well as private schools serving middle- to upper-class families (n = 4). Four teachers taught in half-day programmes and all but one teacher reported using the Creative Curriculum (the other teacher implemented a school-created curriculum). Two of the study teachers were male and the rest were female. Study teachers had a range of teaching experience, from 2 to 20 years, and a variety of educational backgrounds, with five holding a Child Development Associates degree as their highest degree and the rest holding an early childhood-related bachelor’s degree as their highest degree.
Observers
Two different participant groups served as external viewers, to observe instruction. The first group involved observing teachers, both of whom were master teachers at a university-based laboratory school. The observing teachers had 3 and 10 years of experience teaching in early childhood and held a bachelor’s degree and a master’s degree, respectively, in education-related fields. The second group involved observing researchers. These researchers were the first and third authors of this study. Both held PhDs in education-related fields and specialized in research on literacy instruction in early childhood.
Data collection
Study teachers were invited to participate in research about their literacy instruction. After agreeing to participate, they were observed and video-recorded twice during their morning instruction, which included literacy instruction. After each observation, the first author identified two instructional episodes that focused on literacy, ‘instances’ of literacy instruction. These were practices that would be observed in common literacy observational measures (e.g. CLASS, ELLCO, ISI) and that have the potential to develop children’s skills in the areas of: reading comprehension, vocabulary, oral language, alphabet knowledge, print awareness and phonological awareness (e.g. National Early Literacy Panel, 2008; Snow et al., 1998). During observations, the first author recorded the exact times on the video for which this instruction occurred. After the observations, study teachers then participated in stimulated-recall interviews about those instructional episodes (Clark and Yinger, 1997; Gass and Mackey, 2000; Shavelson and Stern, 1981), in which they were asked, ‘In that moment, what were you thinking?’ followed by, ‘why would you think that/or why would you focus on that?’ This process was enacted twice in each interview for each teacher, resulting in a total of 45 episodes of instruction (one teacher withdrew from the study after the first observation and, due to a camera malfunction, study-teacher data from one instructional episode were not available).
After a brief interview about their background experience, all observers (except the first author-observer who completed this prior to the stimulated recall interview) viewed the videos of the 45 instructional episodes. Observers could view the instructional episode as many times as necessary, but only within the time markers viewed by the study teachers. Observers then typed their responses to three questions about the instructional episode: (1) What do you think the teacher is doing? (2) Why do you think the teacher is doing that? and (3) What literacy skills is the teacher targeting? Each observer completed these questions for all 45 instructional episodes.
Data analysis
All study-teacher interviews were transcribed for data analysis. Transcripts were organized by instructional episode and combined with each observer’s responses to the instructional episode, such that data points from all participants were displayed together. Given our research questions, we did not use the videos to observe how children experienced practice; rather, we focused on stakeholders’ perceptions of that instruction. Thus, we analysed the data with respect to how observers and study teachers perceived instructional episodes and how their perceptions compared to each other.
We used a thematic qualitative approach (Braun and Clarke, 2006) to analyse the data. The goal of thematic analysis is to find patterns in qualitative data by becoming familiar with the data and then identifying, describing and naming codes (Vaismoradi et al., 2013). Codes are then used to map or understand narrative data, provide evidence for emerging themes and provide rich descriptions of participants’ perspectives (Creswell, 2007).
We conducted three phases of thematic analysis. In the first phase, each instructional episode was read independently by the first two authors who memoed about emerging patterns and discussed initial findings. Both coders noted differences in the language that participants used in response to interview questions (i.e. ranging from academic to colloquial language); however, we decided to focus on understanding participants’ overall perspectives of practice, returning to language differences later.
Alignment by participant type.
Note. Due to rounding, some totals do not equal 100%.
We coded participant responses as full alignment, partial alignment or no alignment. Full alignment occurred when there was agreement about the gist of the instructional episode among participants, partial alignment represented instances in which there was some agreement about the gist of instructional episodes and no alignment indicated that participants did not agree about the gist of instructional episodes. The first two authors coded every instructional episode independently and then met to discuss this coding; disagreements were reconciled through discussion and in consultation with the codebook.
Understanding nuances in alignment
Because we coded the majority of the instructional episodes as demonstrating partial (n = 33) or no alignment (n = 8; only 4 exhibited full alignment), we subsequently conducted a second phase of analysis. In this phase, we further examined those instructional episodes demonstrating partial alignment. Through this analysis two main patterns in participant differences emerged – themes related to the goals for the instructional episode and themes related to the educational context.
Themes related to goals described what participants perceived as the intended purpose of the instruction. For instance, during an episode where a study teacher worked with children on planting flowers, she used the word germinate. She commented that, ‘I just wanted them to understand what germinate means.’ Her goal was to teach the children a new ‘big word’. This participant’s response to questions about the instructional episode evidenced her intent to teach vocabulary. We compared differences among participant types with regard to the goals they outlined to determine whether there were patterns in those data. Importantly, in all of the episodes where there was no alignment between observers and study teachers, this was due to differences in perceptions of the instructional goal.
We observed themes related to context only in the study teachers’ reports. These themes were related to factors in the classroom that influenced the study teachers’ thinking about instruction. For example, one study teacher commented, ‘I think because we have mixed age, three to five, some of our kids are ready for things that some of our other kids aren’t.’ This quote reflects the educational context within which the teacher operated and enacted instruction. Observing researchers and observing teachers did not report on how educational context informed what they observed about instruction.
Additional coding
First, I was surprised that she knew the ‘W’ because last week, week before, she did not know the W. So, um, that was exciting. Um, and then I just—after that, I just wanted her to know what sound it did make.
Definitions and examples of themes and sub-themes.
Note. ST: study teacher; OT: observing teacher.
Table 2 defines the categories with examples of sub-categories for context and goals.
Findings
The purpose of this study was to understand how study teachers, observing teachers and observing researchers perceived literacy instruction and the extent to which those perceptions aligned. We observed that, most of the time, participants noted the literacy instruction in observational episodes, but domains other than literacy also emerged in perceptions of instruction for study and observing teachers. Study teachers referenced literacy instruction across 82% of instructional episodes and referenced other content in 35% of instructional episodes. Similarly, observing teachers discussed literacy instruction across 91% of the instructional episodes and referenced additional content areas in 47% of the episodes. Observing researchers discussed literacy instruction in all of the episodes but only once identified additional content areas (<1% of instructional episodes).
When considering alignment across participants, there were few instances of full alignment or no alignment among their perceptions of the instructional episodes. Instead, participants’ perceptions of what they observed, what the study teacher intended and what literacy goals were targeted ranged along a continuum of alignment to misalignment. In other words, although participant types often agreed about the gist of instructional episodes, they also contributed unique perspectives that sometimes differed or added more nuances to understanding the instructional episode. More specifically, study teachers added information about contextual factors that were part of their instruction, such as attending to multiple instructional goals simultaneously, whereas observing teachers and observing researchers did not. Interpretations of instruction also seemed to be tied to the background with study teachers using their in-the-moment experience of working with children, observing teachers examining instruction through a practice lens based on their teaching experiences and observing researchers interpreting instruction through an academic lens. These differing perspectives tended to centre around the identification of additional goals of the instructional episode and the role of context in interpreting the episode. Next, we discuss how these patterns contributed to understanding alignment across perceptions of the instructional episodes.
The continuum from alignment to misalignment
Only four instructional episodes, less than 10% of all episodes, evidenced full alignment across all participant types. For example, in one instructional episode in which the study teacher was showing a child four picture cards and naming the pictures on the cards, observing teachers noted the goal of instruction was to teach ‘letter sounds’, and observing researchers noted the goal was ‘letter-sound correspondence’ and ‘phonics’. The study teacher reported the goal was to ‘give ‘em the sound and help put the words together’. The perceptions were the same across participant types in this example, showing full alignment.
Full alignment within participant type was more common, such as between observing teachers who agreed about what happened during instructional episodes 44% of the time, and between observing researchers who agreed with each other the most, or 76% of the time. Observing researchers agreed with study teachers (27%) more than observing teachers agreed with study teachers (18%), and more than observing teachers and observing researchers agreed with each other (22%).
Complete misalignment occurred when participants’ perceptions did not match at all. Similar to full alignment, complete misalignment was not typical among participants. There were only eight instructional episodes (18%) in which there was complete misalignment. Differences among participants in these episodes were solely about the goals of instruction, such as during one episode in which a teacher used a calendar with his children during group time. The observing teachers reported that the goal was ‘numbers and letters in name’ and ‘awareness of text-numbers’, observing researchers reported that the goals were ‘name writing’ and ‘alphabet knowledge’ and the study teacher reported the goal was to keep track of which numbers had been created by which children for the classroom calendar. The study teacher’s goal was to keep materials organized. However, observers noted other valuable aspects of the instruction, including developing literacy and numerical skills.
There were differences among participant types with regard to how often they misaligned. Referring to Table 1, observing teachers and study teachers misaligned about what happened during instructional episodes the most (24%), followed by misalignment between observing researchers and study teachers (18%) and misalignment across all participant types (18%). There were also differences in misalignment within participant types, but those differences were smaller than across participant types. Given that there were few instances of either full alignment or complete misalignment across participant types, we focused on instances of partial alignment, in which participants observed commonalities in the instructional episode.
Partial alignment occurred when participant responses were similar, but at least one participant added information about the episode, or perceived something about the episode that differed from the rest. Across all participant types, 73% of instructional episodes were coded as partial alignment. Importantly, instances of partial alignment generally reflected a common literacy focus among participants while also including other elements of instruction. Unlike misaligned instances wherein participants disagreed about the literacy focus. For example, in one instructional episode, the study teacher was working with children at a smart board with letters and each child had to take their turn using the letters. Observing teachers reported that the goal was to teach ‘sequential order, letter identification’ and ‘social emotional, letter identification’ skills, whereas observing researchers reported that the goal was ‘letter recognition’ and ‘alphabet knowledge’. This example illustrates how participants partially align, agreeing about the literacy aspects of the instruction but differing about the other elements perceived as embedded within the instruction. It also suggests that participants’ perceptions might relate to their educational or experiential backgrounds with the use of more academic language by the observing researchers and a focus on social-emotional development by the observing teachers. In addition to differences in participant perceptions’ of instructional goals, such as the previous example, we also found instances of divergence with regard to context-related elements present in study teachers’ descriptions of their thinking. Next, we discuss differences related to instructional goals and context for the 41 instructional episodes that evidenced partial or no alignment.
Goals
Instructional goals related to literacy were the focus of this study. Observing participants were directly asked what literacy skills were targeted in the instructional episodes they observed and study teachers were informed that the focus was on their literacy instruction. The questions posed to the observers probably shaped how they reported their perceptions of the instructional episode. Specifically, observing researchers reported the perceived goals of instruction as literacy in all of the observed episodes. Despite this focus, the other participant groups, comprised of teachers, identified additional content areas that may have been more salient to them, as was the case in an instructional episode in which the study teacher told a child, ‘… we are trying to protect our clothes. That’s why we are wearing our smocks.' The observing teachers commented that the teacher was ‘managing a child’s behaviour’ and ‘trying to keep the child’s clothes clean’. In this case, the observing teachers diverged from the literacy prompt to report the goal was ‘self-care’ and ‘social emotional’. Similarly, the study teacher reported that, ‘some of our parents don't really care when their kids get messy, and some of them do.' She wanted to respect parents’ rules about keeping clothes clean. The observing researchers, however, followed the research protocol and solely identified the literacy content in the instruction as vocabulary instruction, helping the child identify what the word protect means. Background may have influenced this response pattern as the researchers focused on the research question, and the teachers reported on the most salient feature of the instructional episode, given their classroom experiences of messy clothes and parental concerns.
In addition to the literacy focus of instruction, study teachers reported other goals of their instruction (see Table 2), including goals related to: social emotional development; kindergarten readiness; assessing learning; science, technology, engineering and mathematics (STEM) goals; and classroom management. At times, study teachers discussed multiple goals for one instructional episode. The most frequently reported other instructional goals were related to assessing children’s learning and kindergarten readiness, and these goals were only discussed in relation to instruction by the study teachers (n = 14 and n = 8, respectively). An example of an instructional goal related to kindergarten readiness was when a teacher asked a child what the beginning sound of a word was. She reported she did that because, ‘in kindergarten they need to sound things out in order to be successful.'
The next most frequent non-literacy related goal reported was developing social emotional skills. Importantly, this goal was perceived across 17 instructional episodes by both study teachers (n = 8) and observing teachers (n = 13), but not by observing researchers. Indeed, observing teachers reported a social emotional-related goal in seven instructional episodes where study teachers did not identify this as an intended goal. Sometimes these social-emotional goals co-occurred with a literacy goal, as in the instance in which a study teacher was defining the word ‘nervous’. In explaining her reasoning, she commented that she taught the vocabulary word to help the children express themselves, stating ‘because if you can’t express yourself, you get frustrated’. In this instance, both the study teacher and one observing teacher discussed the social emotional value of the instructional episode. Interestingly, we frequently noticed this pattern across the teacher–participants’ responses. It seemed that often the social emotional aspects of learning were not divorced from academic goals for instruction.
Less frequently reported instructional goals by study teachers were STEM-related or efforts to assess a child’s learning, each reported six times. Goals that were perceived as STEM-focused were reported by observing teachers or study teachers the most (n = 6), whereas observing researchers identified a STEM goal just once. STEM-focused goals largely referred to goals such as teaching counting, sequencing, colours and natural science concepts, such as when a teacher introduced the work ‘germinate’ to expand the children’s vocabulary and also to teach them about how flowers grow. Goals that related to checking a child’s learning were reported six times, and only by study teachers. For example, when a teacher was working on rhyming with his class, he noted that, ‘… I wanted to know if they heard it and listened to it.’
Finally, goals that related to classroom management were reported in three instructional episodes, and only by study teachers. For instance, one teacher discussed her thinking about when to engage students in a particular exercise in order to keep her classroom managed. She stated, ‘I had it in my mind that they were going to throw the snow, but then I had to rethink: this is morning. Morning is a lot more calm than the afternoon.' This teacher connected her instructional decisions to her strategy for classroom management and to the context of her classroom environment.
Context
Differences in context were observed when participants referenced a contextual element in the instructional episode that included knowledge of one specific child, children in the classroom, the classroom environment, classroom routines and curriculum (see Table 2). Importantly, references to these contextual factors were only added by study teachers and not by observers, as observers did not have the embedded understandings that study teachers did regarding their teaching contexts. Most of the instances of context related to a study teacher’s knowledge of a specific child, or children, in their classroom (n = 25). One study teacher reported that she introduced a new word to a particular child in her classroom because, ‘it’s not like, you know, when you tell kids a new definition and they-they never visit it again. Like, so, Jen will use it again.' Here, the study teacher used a new word that she thought Jen would understand and use. Her knowledge of Jen’s educational progress contributed to her decisions about instruction in this particular episode.
In addition to discussing how individual children’s needs related to their instruction, teachers often (n = 12) referenced how multiple children in the classroom factored into their instruction. That is, how they considered several children simultaneously as they taught. One study teacher reported her instructional goal, to teach rhyming words, was led by children’s interest in a particular story. She reported that ‘I felt like they were very interested in the story, so I’m just pulling out anything I can from that story, and rhyming was just one of the things from the story.’ She elaborated that the children in her classroom enjoyed that book, and she used it to teach an academic concept. This example illustrates teachers’ contextual knowledge of classroom dynamics, and it connects to the next contextual element reported by study teachers: curriculum.
Curriculum emerged as a sub-theme in study teachers’ perceptions of their instruction 15 times. For instance, a study teacher described relating classroom activities and children’s interests to literacy goals. Specifically, she reported that children in her classroom were interested in riddles. She took their interest and built literacy-related activities around books that included riddles. She stated that a riddle book ‘helps with … letter recognition, sounding out letters, and that’s—those are the things we’ve been workin’ on.’
Knowledge of specific children, knowledge of children within a classroom and curriculum contributed the most to the contextual elements that we identified among study teachers’ responses. Importantly, many of these themes seemed to emerge in tandem during study teachers’ discussions of their thinking during an instructional episode. That is, study teachers discussed how multiple contextual variables informed their thinking during instruction in a way that observers did not.
Differences in ‘labelling’ or naming literacy skills
One difficulty in examining these data for alignment was the differences in naming/labelling the literacy skills targeted. Specifically, some participants provided a very narrow description of a skill compared to other participants who described the skill more broadly. For example, when describing the targeted literacy skills in one instructional episode in which the study teacher responded to a child’s use of the word ‘childs’ instead of ‘children’, an observing teacher wrote that the literacy skill being targeted was ‘irregular plural’, whereas an observing researcher described the focus as ‘syntax’. The observing teacher focused on a very specific skill, and the researcher on how this skill fits into a broader domain of literacy skills. This example also highlights differences in language that one might anticipate based on the backgrounds of the participants. Syntax might be a more academic or ‘jargon’ word expected to be used by a researcher and not necessarily an early childhood teacher. Indeed, when the study teacher described her thinking about the instructional episode, she said that she wanted the child to get the ‘right grammar, proper language’, not discussing syntax or plural forms of words.
Another example of this contrast in the way different participant types labelled a literacy skill involved an episode in which the study teacher had a paper fan and asked the children to write the word fan. The observing teachers described the targeted literacy skills as ‘connecting visual representations with written words’, ‘emergent writing’ and ‘letter sounds’, whereas observing researchers described the skills as ‘print concepts’ and ‘alphabet knowledge’. The study teacher reported her goal was for the child to write the letter ‘F’ using a reference to the letter F on an alphabet chart. All participant types agreed that the study teacher targeted the child’s writing, but the language they used to describe what the teacher did varied. This observation parallels other findings that participants’ responses tended to reflect their various roles in education. Although responses employing different terminology but encapsulating the same concepts were coded as being in alignment, this finding regarding terminology differences was considered relevant to understanding how perceptions align across teachers and external viewers.
Discussion
This study examined how perceptions of observed instruction aligned across a variety of stakeholders, finding that participants’ observations of instructional episodes aligned along a continuum of agreement regarding instruction. Importantly, there was little complete misalignment across perceptions of literacy-related instructional goals, yet, at the same time, there was rare complete alignment across perceptions of overall instruction. Often these perception differences were due to an incomplete understanding of study teachers’ intentions and external observers tended to omit non-literacy goals and contextual elements important in study teachers’ thinking about instruction. These findings demonstrate the complexity of classroom processes and underscore the multiple perspectives that can be brought to bear on the same instance of instruction. Next, we discuss these findings and their implications for practice.
The importance of alignment for understanding instruction
In general, we observed little complete misalignment between study teachers and those observing their practices, with observers and study teachers only completely disagreeing in 18% of the episodes of instruction. This is a promising finding as it suggests that observations conducted by external viewers can capture at least part of the intended practice of teachers. Additionally, when full misalignment between observers and study teachers occurred, this often involved disagreements regarding the goals of the activity. In these cases, the observers were likely to see the potential additional literacy value that teachers were achieving with their actions, attributing literacy goals to practice not intended by the teachers and giving credit to the teacher for engaging in literacy instruction even when there was another goal in mind. Thus, it seems that observations may also have the advantage of identifying practices that the teacher may not know or have identified independently. Together, these findings add to the existing literature and suggest that observations and observational measures, which are so critical to the field, may be able to capture at least part of teachers’ literacy-related intentions, as well as identify additional practices, thus supporting their continued use in early childhood settings (Connors and Morris, 2015).
The importance of observer background for understanding instruction
How observers perceived instruction seemed to be tied to their background experiences. All participants were asked about the literacy components of instruction, but observers and study teachers referenced domains beyond literacy. These differences fit in the middle of the alignment continuum, where participants evidenced some commonality, but also added their unique frameworks to observations. For instance, observing teachers discussed the social emotional aspects of instruction, while study teachers added information about their focus on kindergarten readiness along with how their knowledge of specific children or a child within their classroom contributed to their thinking about instruction. Observing teachers’ experiences in their own classrooms probably coloured their responses with these teachers, tending to identify instructional goals beyond literacy, in particular social emotional development-related goals. These teachers were incorporating their own instructional experiences and probably their reflections on their own thinking during instruction as they observed the study teachers. In contrast, researchers focused on the literacy value of the instruction, attending to the specific research prompts used in the study. Participants’ backgrounds played a key role in what they identified as salient about classroom instruction.
Thus, it may be that partial alignment, or the middle of the alignment continuum, reflected the practice- or research-oriented perspectives of participants. Sometimes, these perspectives aligned in the way literacy instruction was perceived, but sometimes, they differed in a way that matched participants’ frameworks. This is important when considering who uses observation and for what purposes. As mentioned in the literature review, teachers may be observed for a variety of purposes (e.g. professional development, research, quality rating systems) by a variety of people (researchers, supervisors, peers); both the purpose and the person shape the observation and the interpretation of practice. This study highlights the importance of recognizing how these orientations shape how viewers see instruction. Indeed our data suggest that more research is needed to fully understand how whom the observer is shapes the interpretation of instruction.
Another important difference that emerged across observer type was the ways that they described literacy practices. Although participants were often discussing the same intent, the ways that they languaged their perspectives differed. In other words, participants were often describing similar literacy-related constructs but either at differing grain sizes of the construct or using differing terminology (e.g. ‘alphabet knowledge’ vs ‘letter learning’). It is possible to see how such variations relate to their experiences and academic or practice preparation. The researchers’ language use reflected that used by experts with terminal degrees, whereas teachers’ language was less academic reflecting their range of educational backgrounds. These language differences are important as they make it challenging for a variety of stakeholders to communicate with one another. Specifically, in cases of evaluation or professional development, using differing language for the same constructs could lead to confusion regarding the topic at hand. Thus, this study underscores the need for the field to think about the ways of communicating about literacy instruction such that all stakeholders are describing the same constructs.
The importance of continuing to focus on the ‘hidden side’ of teaching
Our findings suggest that, although they may differ in how they language their perceptions, external observers can generally identify the same literacy goal but do not necessarily understand all of the intent, context or additional instructional goals of the teacher enacting the practice. This suggests that standardized observations and external observations are accurate at a certain level and for a certain purpose – they are just not capturing the full picture of practice. This was evidenced in the majority of partially aligned instructional episodes where teachers’ background knowledge of their classrooms and children was not accessible to observers, yet it was still part of teachers’ instructional enactment. Other studies have also identified the role that contextual knowledge plays in informing teachers’ instruction (Lampert, 2003; Shulman, 1987). This study illuminates the gap between observation and the ‘hidden-side of teaching’, which may be missed by strategies that address observable features of classroom processes. Thus, observation and observational tools only provide one picture of instruction, omitting the key role of intention in teachers’ enactment of practice.
Understanding teachers’ intentions and thinking is critical when considering mechanisms for improving literacy-related instruction. Understanding what teachers are thinking about when implementing specific literacy practices provides insights into how teachers bring these practices about. In other words, considering the ‘hidden side’ of the work, via teacher intention and thinking, provides context for why teachers make moment-to-moment decisions in the classroom, which ultimately lead to enacted practice (Schachter, 2017). Conversely, understanding teachers’ perspectives of practice may also illuminate why teachers choose not to implement certain literacy practices in the classroom and thus explain why these practices are not observed as frequently in classrooms (e.g. Justice et al., 2008; Pelatti et al., 2014). For teachers, there is more to consider than whether or not to simply engage in a literacy-related practice. For example, there are the needs of the children as well as other instructional goals to consider.
This is important to bear in mind as many researchers use literacy-related observation measures to provide professional development (e.g. Al Otaiba et al., 2016; Cabell et al., 2013; Landry et al., 2011). This study, along with other professional development literature (e.g. Desimone, 2009; Schachter, 2015), indicates that simply structuring professional development to teach specific literacy-practices is not enough to support changes in practice without attending to teachers and their views of instruction. It may be necessary in professional development contexts to support teachers in thinking holistically, incorporating contextual factors such as curriculum and additional goals into their discussions with teachers as they support them to use these practices (Flynn and Schachter, 2017; Shafto et al., 2014). Combining outside expertise provided by researchers, coaches or directors with the situated knowledge of teachers may be the key to providing professional development to support the use of literacy practices that best support the needs of individual children.
Finally, these findings suggest that for those seeking to understand the complexity of classroom instruction, observational-based measures should be used in tandem with other measures that allow deeper insights into teachers’ thinking about practice. For example, the ELLCO includes a teacher interview that provides more information about teachers’ thinking regarding literacy instruction and, if used, could be beneficial for learning more about teachers’ intentions. Additional research into strategies that combine both observational data along with teacher-perspective data is needed.
Limitations
There are some limitations in this descriptive study. Both the first and the third authors of the article were participants in the research study and the first author both collected and analysed the data. It is probable that these experiences informed the authors’ interpretation of the data. However, to reduce possible bias, the second author was intentionally included for her neutral position in relation to the data. There was a small pool of stakeholders sampled in this study (researchers and classroom teachers), which may not reflect other ways that observers perceive literacy instruction. As we have argued previously, more research is needed to understand how a variety of observer backgrounds and roles shape our understanding of instruction.
Conclusion
This study described the nuanced differences between what external viewers observed in teachers’ instruction and teachers’ own perceptions of that same instruction. Importantly, we found common ground among participants regarding the purpose of instruction most of the time. However, there were interesting variations in participants’ perceptions that highlighted the complexity of classroom processes, the value of teachers’ contextual knowledge and the multiple perspectives brought to bear on the same instance of instruction. This study highlights that observation and subsequent tools can capture aspects of classroom practice, yet may not fully encompass the important ‘hidden side’ of teachers’ work. Furthermore, it suggests the importance of incorporating context and teacher thinking into professional development efforts aimed at improving literacy instruction.
Footnotes
Acknowledgements
The authors would like to thank all of the teachers who participated in this study without whom these important insights into practice would not be visible.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work reported here was supported by the Spencer Foundation.