Data Collection
Decoding Survey
The first method of data collection chosen was a diagnostic decoding survey that served as a pretest and posttest. This method was selected because it provided an initial breakdown of how all three students were able to decode words. On the survey, there were 50 words in total, a point given for each word the student was able to pronounce correctly. By third grade, the average amount of words that should be able to be pronounced correctly is 48. If the student was not able to correctly decode the word, no point was awarded, and further analysis was done to find where the error occurred.
Each word was broken down into its phonemes, for example, consonants, blends, digraphs, and short vowels. A mark was made next to where in the word the mispronunciation happened. After analyzing each word, I calculated which sound had the most errors. The results of this survey informed my decisions on where each student began in terms of specific phonemes. For all three of my students, the majority of decoding errors were in short vowel sounds and consonant digraphs. I decided to start my instruction with short vowel sounds and built towards consonant digraphs. This method gave me baseline data and determined growth from their baseline to their final day of the action plan.
Running Records
The second method of data collection was running records. This method was chosen because it allowed me the opportunity to see how each student used their decoding skills in the context of a text rather than individual words. Once a week, each student read a text at their appropriate level, using the Fountas & Pinnell system, aloud while I gathered data on the accuracy of their reading. The number of errors the student made determined the accuracy score. The fewer the number of mistakes made, the higher the accuracy percentage. Once they were at a 95% or above in accuracy with at least satisfactory comprehension, they moved to a more challenging text. If there were instances where a student's accuracy score was high, but their understanding of the text was limited they stayed at the same level text.
Using running records allowed me to see each week where the individual student was in their ability to decode words. For instance, if errors were still frequent in words using short vowels, that told me they needed additional assistance in that area. It also provided me more accurate data to determine if the student was ready for a more challenging text.
Progress Monitoring
A progress monitoring system was put into place as my third method of data collection. Each student had a binder that held their progress monitoring tasks and their continued progress data that both my paraprofessional and myself collected. Every week, my paraprofessional and I would track each student based on their ability to decode words. They were given a list of ten words, and once they were able to read all ten words correctly, they moved onto more challenging material. All three students started by decoding short vowel-consonant words. After the student was able to decode at 100% accuracy, he or she moved to consonant-vowel-consonant words which are slightly more challenging. Then the process repeated, except they graduated to words using consonant digraphs. This system gave them the opportunity to interact directly with phonemes that were proven challenging, based on the results of the diagnostic survey. Using this system helped ensure that students moved at an appropriate pace and focused on material that challenged them but did not overwhelm them.
Anecdotal Notes
Additionally, anecdotal notes were taken throughout the study to provide qualitative data. These notes were used to monitor how the study implementations affected students behaviorally. All three students were known to display off-task behaviors in different ways when given a challenging task before the study. Further, one of the students was irregularly taking their prescribed medication which led to off-task behaviors. If lack of focus had been a consistent challenge throughout the week that would change the daily activity. I switched from a more challenging activity, like writing sentences, to a more relaxed exercise that all three responded well to, like sand writing. Subsequently, I also documented the number of redirections needed on average to see if there was a difference between the beginning of the study and its conclusion.
Data Analysis
Pretest and Posttest
This graph illustrates each student’s individual growth in decoding from the beginning of the study to the end by using the same decoding survey as a pretest and a posttest. This survey included a list of 50 words using a variety of phonemes and graphemes. Every word they were able to pronounce accurately counted as a correct response. By third grade, a student on average should be able to read 48 out of 50 words correctly, represented by the purple line on the graph. When given this survey as a pretest, all three of my students scored significantly below average. This data served as a benchmark in deciding where my students needed explicit phonics instruction. Retaking this survey in the spring showed significant improvements. The orange bars on the graph were the results of the pretest, and the teal bars were the results of the posttest. Student A showed the most growth with a 121% increase, Student B held a 44% increase, and Student C grew at a 65% increase. All students showed significant growth. As a result of the posttest, Students A and C were now close to reaching the average score for a third-grade student.
Short Vowels and Consonant Digraphs
When analyzing the results of the pretest, patterns began to emerge from each student's mispronunciations. The most common errors for all three students were in words that included short vowel sounds and consonant digraphs. On the survey, 26 out of the 50 words used short vowels and eleven included consonant digraphs. I split this data into two graphs, one for short vowels and the other for consonant digraphs. First, I recorded how many of those words they pronounced correctly on the pretest, represented by the orange bars on both graphs. Then, I calculated how many they were able to pronounce correctly on the posttest, depicted by the teal bars. For example, on the pretest Student C pronounced eleven out of the 26 short vowel sounds accurately. On the posttest, Student C pronounced 24 correctly. When I compared the results from the pretest and the posttest, there was growth for all three students.
With short vowels, Student A had a 425% increase in decoding accuracy, Student B had a 20% increase, and Student C had a 100% increase. Student B scored the highest on the pretest in short vowels. This high pretest score could have contributed to the smaller percent increase because there was less area of growth needed. Student B needed more direct instruction in digraphs rather than vowels.
For consonant digraphs, Student A had an 800% increase, Student B had a 75% increase, and Student C had a 25% increase. Student C scored the highest on the pretest in consonant digraphs. This prior knowledge in consonant digraphs could have contributed to a smaller percentage increase due to a smaller area of growth needed. Opposite of Student B, Student C needed more assistance in short vowels.
Reflection and Triangulation
My lingering question was that Student A and Student B started the study at a very comparable place. What caused the dramatic increase of Student A compared to the smaller growth of Student B? Student B was on an irregular medication schedule that was prescribed to aid in focus. On days when the medication was not taken, participation in reading activities was a struggle. This inconsistent schedule could have contributed to the smaller and uneven progress shown in the progress monitoring and daily redirections data. On the days Student B struggled to focus, more of my attention went to Student's A and C which again could have contributed to the amount of growth they achieved with more of my direct instruction. Student A showed tremendous growth between the pretest and the posttest, but that is not mirrored as significantly in the guided reading progress. To me, this indicated that using these decoding skills in the context of a text needed more emphasis. An improved ability to decode a word was an improvement, but the end goal was to use those skills in reading.
Running Records
At the beginning of the study, each student was benchmarked using the Fountas & Pinnell reading system to determine which level of text was appropriate. Once a week, each of the three students read aloud to me while I completed a running record. The majority of the running records were cold reads, meaning the student did not practice the reading prior. My reasoning for doing so was to ensure that I had accurate data on how the student decoded independently. They were monitored using an overall accuracy score of how they were able to decode the text. Once they received a 95% or above in accuracy with at least satisfactory comprehension, they moved to a more challenging text. The graph to the right tracked the progress of all three students through the six-week study.
Each Student's Guided Reading Growth
Student A was benchmarked at a level B and finished at a low E.
Student B started reading at a level B and finished at a low D.
Student C was benchmarked at a C and ended at an E.
Each student made improvements in their reading by moving at least two levels, moving them from kindergarten level reading to the beginning of first grade. Student A made the most progress by moving four levels. In the data, for all three students, after they reached a high-level accuracy, there was a sharp dip in the graph. This dip was caused by them testing out of an easier text and moving to a more difficult text. With navigating a more challenging text came decoding more challenging words, this could have resulted in the accuracy percentage falling. After the dip, for all three students, there was a slow incline back to a higher accuracy level as confidence increased. An outside factor that could have affected guided reading growth was time spent reading outside of school. Students were asked to read fifteen minutes a night to an adult. A parent or guardian was asked to sign a log as proof that they read. All three students were inconsistent; however, Student C read the most frequently. A strict schedule of reading every night could have had a positive effect on their growth.
Reflection and Triangulation
Students B and C showed a sharp decline in their fourth dot representing week four. This drop could be attributed to behavioral issues seen throughout week four with a decrease in progress monitoring accuracy and a higher amount of redirections needed that week shown in the graphs below. With the majority of the behavioral issues stemming from Student B, my lingering question was how much did that affect not only Student B’s results but Student A and Student C? Students A and C had lower accuracy in running records and progress monitoring in week four along with more redirections needed. This data proved that there was an impact on success for the week which could have detracted from their overall achievement. Student B's running record accuracy was lower on average than Students A and C. I looked at my other data to find possible reasonings for this pattern. The day of the week that Student B averaged the most redirections was on Tuesday. Tuesday was also the day that I had set aside for Student B to complete a running record each week. One possible explanation for the lower accuracy was that running records caused off-task behavior due to anxiety. Another possible reason for the low accuracy score was that Tuesday was the most likely day for medication to be missed interfering with the ability to concentrate on reading.
Progress Monitoring
This graph reflected the results of progress monitoring done throughout the study. It is mapped by accuracy score. For instance, if they read four words out of ten correctly, they received an accuracy score of 40%. Once those ten words were read correctly; they moved to a more challenging set of words. All three students started with vowel-consonant (VC) words and moved to consonant-vowel-consonant (CVC) words and finally to consonant digraph (CD) words.
Similarly to running records, once a student reached a high level of accuracy, there was a sharp dip in their growth. This dip could be attributed to moving to a more challenging set of words. Student A moved from VC words to CD words in the six weeks, reaching all three levels of words. Student B went from VC to CVC words reaching two levels of words. Lastly, Student C went from VC words and at week six passed CVC words. Each student showed growth, with Students A and C making slightly more progress than Student B by moving on to CD words. This growth illustrated a dip in week four where behaviors were more prevalent throughout the week. Something that I wondered was if Student B initially scored higher in consonant digraphs than short vowels would it have been more valuable for that student to start with CD words.
Reflection and Triangulation
A pattern throughout the data was that Student C started with the highest accuracy, for the decoding survey, progress monitoring, and guided reading level, in comparison to the other two students. However, Student C also had a slower growth rate compared to the other two students. This pace could have been due to a smaller need, but other factors might have contributed to this as well. Student C was the student who refused to read most frequently when faced with a challenging task or when upset. The amount of instruction time lost possibly contributed to the slower rate of growth.
Anecdotal Notes
Something that I decided to implement in my anecdotal notes was how many redirections were needed on average for all three students. Each day I wrote down how many redirections were required for each student. Then, at the end of the week, I averaged them together to get an overall number of redirections per week. For example, Student A had an average number of five redirections each day for week one. Looking at the data as a whole, the number of redirections needed from the beginning of the study to the end decreased for Students A and B. Student C required fewer redirections throughout the study compared to Student A and Student B. Week four had the highest amount of redirections for all three students.
Reflection and Triangulation
Looking back in my notes I also noticed that weeks three and four were the weeks where Student B was not taking medication. Then, week five when back on medication, the daily number of redirections decreased for all three students. This data showed me the amount of influence Student B’s behavior had on the other two students. Also, when I looked back at my progress monitoring and running records, some of the lowest accuracy scores occurred during those two weeks. Something I wonder was how much more progress could Student B have achieved if taking medication would have been consistent.