Un-quantifiable

My job is great. I wake up excited and challenged: how can I teach my students, all learning English as another language, as much as possible during the coming day? They need, in varying degrees, extensive vocabulary, instruction in English syntax, exposure to English written and spoken text, practice writing and speaking, background knowledge in academic subjects, support in study skills, and perhaps even a basic literacy, which is sometimes lacking in their first language. There is so much to do every day, and I almost feel like new techniques, research, and ideas have been crammed down my throat since I entered this field. There is no question that all these ideas and creative ways of helping kids learn is exciting and keeps things interesting--but I often finish a class and ask myself, "Was that really the best way to do that? Would X, Y, or Z work better?" And the frustrating part is that I'll never know: education being what it is, I will never have the same number of students from the same backgrounds with the same interests and the same proficiency. There is no way to control for all of that, and so truly quantifying what works best is elusive. On occasion I have a hunch that I should have done something differently, or sometimes I'll think what I tried worked out just great, but beyond my intuition, I have little to go on.

For example, I have read repeated reports that reading aloud to students is beneficial. But it takes a significant chunk of class time, and if I also implement independent reading, paired reading, reading strategies instruction, process writing, writing conferences, various models of vocabulary instruction, group presentations, and critical-thinking seminars, how can I possibly make it all fit together in any way that does not seem totally frenetic? I do what I can, but I am constantly wondering if the activities are in the right proportions and presented at the right moment.

At times I think perhaps the answer is meta-analysis: taking huge numbers of data from schools teaching every imaginable type of student, dumping the data into a computer with advanced logarithms (can you tell I have no stats background and have no clue what I'm talking about?) and then coming up with incontrovertible evidence about certain practices that work best based on the sheer breadth of the data.

One of the most compelling arguments I have ever had presented to me in education was a huge meta-analysis of studies investigating the best method of instructing English Language Learners. The researchers compiled every study they could find that met certain conditions of size and quality, then did a statistical analysis on the data to conclude, quite powerfully in my mind, that a program promoting a true bilingual education--that is, instruction in students' L1 and L2s--was always more effective than English only immersion. I was astounded: if this was the case, WHY is this not more widely shared? I am amazed at both how the political machinery can cloud the results and how the academic world has done such a poor job of sharing this result with the general public.

Anyway, back to my point: while it would be a slow, tedious process to determine how to analyze all the components of a well-rounded ESOL program in such a way as to find out what works best in which increments, would it be worth it to have more effective instruction? And then I wonder, how boring would THAT be if all my instructional decisions were obliterated and all I had to do was read a script?