How Young Is Too Young for State Testing?
Bubble-in tests for 5-year-olds? That was one teacher's first reaction when she heard Massachusetts was planning to implement a statewide assessment system for kindergartners. Allison S. (kindergarten, Colorado) found the state was considering observational and portfolio assessment practices rather than pencil-and-paper tests—but she was still shaken. As states vie for Race to the Top Early Learning Challenge funds, will some go too far?
Allison asked colleagues on the Teacher Leaders Network to share their thoughts on assessing early learners. Here are some highlights of the lively interchange:
Marsha R. (middle school science, Kansas) was first to respond:
I say "yes" to more assessment for 5-year-olds crafted by an expert teacher who knows his or her students' needs, the community, and the specific culture and conditions of the school. I absolutely say "no" to standardized testing of young children.
Allison explained that assessment is a critical part of her practice:
Every day, I assess each of my students: Do they understand what I am teaching them? Can they apply these new skills or knowledge? If not, what am I going to do about it, immediately? To me the bigger fear is that others will want to determine how and what my assessments will look like. Most of my assessments are done through casual conversation and multiple observations—there is no way I could capture that in a checklist or one-size-fits-all spreadsheet.
Marsha noted that the public may not realize what teachers are already doing to measure student learning:
Teachers should be loud-mouthed about the fact that we're already doing what needs to be done. That standardized testing will not improve the classroom and will, in fact, take away from the important and effective work that expert teachers are already doing. The public needs us to tell them that we're ahead of this idea, to show that we’ve already devised assessments that are more effective than standardized tests.
Sean W. (high school social studies, Colorado) argued there may be value in having more data, especially in addressing the achievement gap:
Maybe this is where a teacher-created assessment bank should come into play. If master teachers are already using assessment strategies that work, this could be a step toward ensuring uniformity across the country and eliminating that gap between the haves and have-nots. A good amount of research shows that younger students can make up the gap more quickly than the big kids. So if we can identify where the struggling students are, then it is more data that we can use to help solve the problem. I am deathly afraid of the day when my son enters kindergarten because there really is not a lot of data for these lower grades. Who really knows if the students are progressing?
Cheryl S. (kindergarten, California) spoke to Sean's concerns:
Sean, I can tell you with absolute certainty that I know exactly how each of my 31 kindergartners are doing in every subject area. … I collect data every day in multiple ways and use the data to make decisions for instruction.
For example, one of the benchmarks for this trimester is being able to count to 10. A game l call "magic number" helps me assess students' progress: Students all stand in a circle, the teacher or a student chooses a "magic number," and we count around the circle from zero to that number. The person who says the magic number sits down. We continue until all are sitting.
A student who is unable to give the next number likely needs to work on counting skills. I use this information to create "focus" groups: Students who understand numbers to 20 are working on solving oral mathematical story problems with numbers up to 20, while students who are still working with numbers to 10 are solving the same problems but with smaller numbers.
Even in kindergarten we have TONS of data, we assess our students formally and informally, and we have multiple paths to meeting the needs of every student.
Susan G. (recently retired, high school, Virginia) has doubts about early childhood assessments that are overly reliant on closed-response items:
During our first visit to my daughter's kindergarten classroom, the teacher did some letter and number recognition activities with Rebecca, but the conversational part was much more interesting. In one activity, Rebecca was shown pictures of a horse, a pig, a duck, and a rabbit, and was asked, "Which one doesn't belong to the group?" The grown-up answer, of course, was "the duck."
Rebecca responded, "The horse." The teacher asked her to explain: "Because he is gray and the others are white," Rebecca explained. (True, on the black-and-white handout, the horse was gray, but would a grownup have considered this?)
"Is there another animal that's not like the others?" prompted the teacher.
"Yes, the pig," Rebecca replied confidently.
"Could you tell me how the pig is different?" asked the teacher.
"It's prickly and the others are soft." Rebecca explained. (Earlier that week, on a pre-school farm field trip, Rebecca had been surprised by the boar bristles.)
The teacher was completely accepting in her reaction to these "different" perceptions. Closed-response testing for very young children tells us very little about what they know. It can only reveal whether or not they have been acculturated to give the answers adults want and expect to hear.
Tiffany B. and John H. indicated that performance-based assessment practices might hold promise—but both highlighted the time and personnel investments that states would need to make.
Tiffany B. (peer mentor teacher, Florida) said:
The year after I moved to teaching another grade, kindergartners started doing math tests with bubbles. ... I think states need to take a step back and find ways to collect the same data that we're using to teach in our classrooms. I HOPE Massachusetts is on to something here, with portfolios to show growth and achievement. Yes, it will be more work for teachers and even more work for those trying to draw conclusions for the state but I think this will take a more accurate temperature of our students. I am very excited to see how this develops.
John H. (preschool teacher, Virginia) added:
Assessment in early childhood is fairly unreliable, especially in the pre-reading years (0 - 7). Kids just don't think abstractly the way that independent assessment requires them to. ... The more we move towards performance based assessment for students and teachers, the closer we will get to the answers we want—but it can be time-consuming for the assessor.
Allison described herself as "hopeful, but hesitant":
Today, during our kindergarten Halloween party, I walked around the room, asking the students to count their M&Ms. The students didn't know it, but I was definitely assessing them, seeing who had mastered one-to-one counting and who needed more help (even when the counting objects were tasty!). My fear remains that there wouldn't be a checklist or assessment that would be open enough to let me assess them in this very unthreatening way—or that even a portfolio might not allow for this type of documentation or sample work. I'm hopeful ... but hesitant.
The kindergarten teachers who took part in this thread argued that they are already assessing young students' learning in meaningful ways, and using those data on a daily basis. And, as several pointed out, traditional standardized testing is not a good fit for young children.
What's your take on measuring early learning? What roles could expert teachers play in ensuring that early-learning assessments are valid and meaningful? How could teachers help the public to understand their collection and use of data?
For additional teacher perspectives on measuring student learning, visit Teaching Ahead: A Roundtable, which focused on the topic this month.