When psychologists want to investigate age differences between groups of children, they frequently use a cross-sectional design (Creasey, 2006; Miller, 2006). In a cross-sectional design different children at different ages are assessed at the same time. For instance, if one were interested in the development of arithmetic abilities, different groups of children at ages 4, 5, 6, and 7 could be given tests that assess addition and the strategies children use to arrive at their answers. In a very brief time the test giver would have an idea of how this important skill changes with age. (If the researcher used a longitudinal design, in which the same children were tested repeatedly over time, it would take them four years to get the same information.) The intent of a cross-sectional design is to allow psychologists to efficiently describe change over time and to identify the various mechanisms associated with those changes.
Cognitive strategies—the deliberate plans children use to organize their problem solving—have been studied extensively in developmental psychology using cross-sectional designs (Bjorklund, 2005). For example, Jane Gaultney (1998) investigated the use of memory strategies in third-, fourth-, and fifth-grade students with and without learning disabilities. Specifically, she examined the degree to which children organized their recall of related items (e.g., sorting and later recalling together items from the same semantic category) and the extent to which such strategy use benefited their memory performance. She found that typically developing children benefited more from the use of strategies than children with learning disabilities. This difference increased with age and was due to the lack of progress by older children with learning disabilities. Her research also showed that children with learning disabilities experienced more benefit when they used strategies to study items for later recall, whereas typically developing children benefited most from strategies when actually recalling the items. Such an approach identifies potentially important age differences in children quickly and provides some insights into the specific problems children of different ages and abilities have.
The use of cross-sectional designs in education implicitly involves a concern for age differences, and this awareness implies an interest in development. A developmental-psychological perspective is concerned with age-related changes and the factors associated with those changes. The deeper issues of how, when, and ultimately why psychological change takes place shape the formulation of research questions, which in turn influence the methods chosen to answer these questions.
While the importance of measuring age-related change should not be underestimated, it is often the normative, or perhaps even more critical, non-normative influences associated with change that most interest developmental and educational psychologists. A broad holistic view of development leads to questions about children's thinking in both social and academic environments that can lead to techniques for enhancing education (Teti, 2006; Bergman et al., 2000).
Several factors must be taken into account in conducting a study using a cross-sectional design. For instance, if researchers are interested in the development of scientific reasoning in school-age children, they can design a study in which groups of children are given problems to solve requiring reasoning associated with (a) developing an hypothesis, (b) deciding how to test the hypothesis, (c) collecting data, and (d) evaluating the hypothesis. Questions immediately surface regarding how they might go about conducting such a study and what pitfalls they must avoid in order to make sure the results of their study are interpretable.
In such a study researchers must ensure that each age and experimental group is balanced in terms of important demographic characteristics. For example, in the study of scientific reasoning, they may give children in grades 3 through 8 different types of science instruction or different textbooks to read. They would need to balance the age and instruction groups for gender. They would need to consider if children from different age groups are from different schools, and if so if the schools differ in any important way, such as socio-economic status (SES) or quality of instruction. Also, they would consider if the children in the various grades or instruction groups differ in academic abilities, motivation, or previous knowledge of the subject matter in any systematic way. Academic and developmental differences are often associated with gender, SES, motivation, and a host of other factors. Researchers may be interested in how these factors affect scientific reasoning, but in order to interpret unambiguously any age or instruction differences they may find in their study, they have to insure that their groups are balanced, as much as reasonably possible, in terms of demographic and related factors.
Another consideration when doing cross-sectional research is the tasks children in each group perform. These tasks should be age-appropriate, so they are not too difficult for the youngest children or too easy for the oldest. Researchers must check that the wording of the problems is comprehensible to children of all ages to be tested. Once they have decided what tasks they are going to give children, they have to decide how to administer them. Children in each group should be tested in a similar fashion. However, because of differences in the social, emotional, and particularly cognitive abilities of children of different ages, it may not be possible to use the identical procedures with all children. What is important is that children of all ages understand the task and what they are supposed to do. As a result, testing procedures may have to be varied to ensure that both younger and older children understand the tasks, but not varied so much as to make the tasks qualitatively different for children of different ages. For example, it is possible to read or rephrase directions, or even use pictorial representations of the directions, to help younger children understand the expectations of the task. Robust results from research are predicated on ensuring that attention is given not only to the design of the study, but also to the manner in which the research design is implemented.
Cross-sectional designs can be used in conjunction with both experimental and correlational studies. Experimental research is the gold standard for answering questions about cause-and-effect relationships. Experimental studies involve the manipulation of one or more factors, or variables, and observation of how these manipulations change the behavior under investigation. For example, Schwenck, Bjorklund, and Schneider (2007) were interested in factors that influence memory strategy development in groups of first- and third-grade children. To assess these factors, they gave some children instruction in strategy use on some trials, whereas others received no such instruction. They also varied the type of materials that children were asked to remember, including items that were either typical of their category (e.g., shirt, dress, pants for clothes) or atypical of their category (e.g., (e.g., socks, tie, belt for clothes). They found age differences not only in how well children of different ages remember, but also in children's abilities to benefit from strategy training. For instance, older compared to younger children required less explicit prompting to use a strategy, were more likely to use a strategy when atypical category items served as stimuli, and generalized a strategy to new sets of words. They also identified differences in children's tendencies to use a strategy but not to experience any benefit in recall, termed a utilization deficiency, something important to educators concerned with teaching strategies in the classroom.
In correlational studies researchers make comparisons between two or more variables. Keep in mind that identifying a relationship between two variables does not imply causation, but if a relationship does exist, further understanding may be gained by pinpointing the nature of the relationship and then conducting experimental studies. Cross-sectional designs have been used in correlational studies leading to understanding about different developmental trends in the use of reading strategies for children with different levels or styles of learning aptitudes (see Gaultney, 1998). For example, in one study (Siegel & Ryan, 1988), researchers compared the phonological processing (specifically, reading pronounceable pseudowords such as blurt) of 7- to 14-year-old children with and without reading disabilities. Both groups of children showed marked improvement in pseudoword reading with age, although at each age tested, children with reading disabilities performed significantly worse than typically developing children. In fact, the pseudoword reading of 13- to 14-year-old reading-disabled children was comparable to that of their 7- to 8-year-old typically developing peers.
A prominent myth exists that practitioners, most notably teachers, have no need for an understanding of research. The poor transfer of scientific research findings to application in the classroom creates a void between empirical inquiry and classroom-based practices. In short, teachers must be able to understand the meaning of research findings and how to incorporate the results into pedagogical practices. This transfer of knowledge poses a challenge for many teachers who may need more than simple exposure to the research literature in order to make fundamental change. The challenge to teach essential knowledge and foster skill development in children is the primary responsibility of the classroom teacher. It is critical that teachers be able to pose and ultimately answer questions about important aspects of the teaching and learning process. Evaluating the outcomes of schooling based solely on students' abilities to engage in particular skills is clearly an inadequate approach to gain an understanding of the entire schooling process. There is no paucity of variables one must take into account when investigating why and how children make academic and social progress, and equally as important are the factors that account for failure to make adequate progress. A holistic view of the full teaching and learning process is necessary to account for the mechanisms underlying learning, and this often involves an ability to interpret the findings of developmental/educational research and sometimes to perform simple studies.
Research is not the enemy of teachers, and it is possible to structure research that can help answer some of the most salient questions about teaching and learning. There is common ground between researchers and teachers. Each group desires to answer questions about children's ability to learn content and develop skills. For many teachers, the daily grind of classroom instruction is reason enough to shy away from any sort of formal inquiry into the teaching and learning process, but it is exactly these daily routines that provide a rich source of potential issues from which to gain further insight. The cross-sectional design is a relatively accessible approach that can be used in the classroom to answer questions about how children change with age in regard to a particular ability or process and how children of different ages respond to different types of instruction. The following paragraphs provide an example of a cross-sectional study that a classroom teacher could conceivably implement, beginning with the formulation of a research question to the execution of the project.
Formal inquiry begins with the generation of questions about instruction and student learning. From here, the challenge is to isolate a particular aspect of the process that warrants attention. Research is most easily facilitated with a clear focus. For example, a teacher may wonder about how students benefit from the use of reading-comprehension strategies to recall information from a science textbook. Perhaps there are age differences in children's ability to benefit from using reading-comprehension strategies that can be traced using a cross-sectional design.
After formulating the research question, it is then necessary to identify groups of children to participate in the study and to determine what, if any, experimental groups teachers want to include. They may decide based on the science curriculum at their school that children much below grade 3 would be too young and children much beyond middle school would be too old for the questions they wish to ask. In the present example, they may include at each grade tested an experimental group that gets special reading-comprehension strategy instruction, as well as a control group that gets standard classroom instruction or perhaps some extra time with the teacher to balance for the amount of teacher-time children in both groups receive. (They may also want to test special groups of children, such as those with learning disabilities, slow readers, or those from less-advantaged homes, and compare them to typically developing children, but the design can be kept simple for this example.) When identifying groups of participants, it is important to recruit approximately equal numbers of boys and girls and to ensure that the different ages and instructional groups are balanced for other factors that may influence children's performance, such as academic achievement and SES.
The next step is to determine the measurement for the study. In this case, the teachers would need to develop or identify a reading-comprehension strategy that they could teach to their participants (likely obtained from the research literature) and a measure to identify whether using such a strategy works, that is, enhances recall of science-text content. They may be able to use tasks already used in the classroom, for instance, questions taken from the science text or teacher-made questions. They want to make sure that the questions are age-appropriate and that the testing methods are similar across ages.
Based on such an experiment, they may find that whereas third-grade children can implement the reading-comprehension strategy, the students tend not to generalize it to new texts and perhaps experience only minimal improvement in recall from its use. In contrast, older children not only use the strategy but also generalize it to new texts and experience substantial recall benefits from its use. Such findings would help in the development of different types of instruction for children of different ages and perhaps cause teachers to ask more questions about what it is about the reading-comprehension instruction that works (or does not work) for children of different ages. Of course, there are a number of possible outcomes from such a study, but the point here is that one is able to get an idea of how strategy use is developed and then make a determination of which strategies to use and when.
Cross-sectional designs permit researchers or teachers to collect quickly information about age changes in some ability. However, in order to assess true developmental change, longitudinal approaches, which test the same children over time, are necessary. Such studies are not out of the question for classroom teachers. For example, the assessment of reading-comprehension strategy instruction on children's recall of science information could be performed once in November and again in May to see if individual children make improvements over the course of the school year.
Research has not always been held in high regard or maintained a prominent place in the typical classroom. This does not have to be the case if research studies are designed to be easily understood and implemented in the classroom. Teachers typically reflect on the teaching and learning process, and the use of research designs simply allows these questions to be tested and better understood.
Bergman, L. R., Cairns, B., Nilsson, L-G., & Nystedt, L. (Eds.). (2000). Developmental Science and the holistic approach. Mahwah, NJ: Erlbaum.
Bjorklund, D. F. (2005). Children's thinking: Cognitive development and individual differences (4th ed.). Belmont, CA: Thomson.
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press.
Creasey, G. L. (2006). Research methods in lifespan development. Boston: Pearson.
Gaultney, J. F. (1998). Utilization deficiencies among children with learning disabilities. Learning and Individual Differences, 10, 13–28.
Miller, S. A. (2006). Developmental research methods (3rd ed.). London: Sage.
Siegel, L. S., & Ryan, E. B. (1988). Development of grammatical-sensitivity, phonological, and short-term memory skills in normally achieving and learning disabled children. Developmental Psychology, 24, 28–37.
Teti, D. M. (Ed.). (2006). Handbook of research methods in developmental science. Malden, MA: Blackwell.