Latest / The Reading Instruction Show / Deconstructing Goldenberg's Science of Reading Propositions

Transcript

Deconstructing Goldenberg's Science of Reading Propositions · The Reading Instruction Show

• 0:01
  This is the reading instruction show on your
• 0:03
  host, Dr. Andy Johnson. This is part two of a
• 0:08
  podcast looking at words and words matter. We
• 0:12
  must hold scholars accountable for the words
• 0:15
  that they use. And in this podcast or show, we're
• 0:19
  going to look at propositions, variables, and
• 0:23
  baloney. Review some common tricks used to create
• 0:27
  convenient inaccuracies, and I've added a couple.
• 0:30
  An academic buzzword is used to create the illusion
• 0:34
  of something important. Buzzwords are used to
• 0:38
  elicit an emotional response rather than accurately
• 0:41
  transmitting information. A cow pie is the term
• 0:46
  used to indicate a pile of academic buzzwords
• 0:49
  wired together to create conveniently inaccurate
• 0:53
  perceptions. Weasel word. Words used to get the
• 0:58
  impression of something without really saying
• 1:01
  something. Straw man argument, creating a cartoon,
• 1:05
  and then arguing against the cartoon. Two new
• 1:10
  words, baloney. Creatively interpreting or misinterpreting,
• 1:16
  research or decontextualize data, and proposition
• 1:21
  -heavy writing. Cramming. an inordinate number
• 1:26
  of propositions into a sentence. These are all
• 1:31
  used to create inaccurate perceptions of things.
• 1:35
  We are focusing on baloney in this part too,
• 1:40
  because baloney harms students, teachers, and
• 1:44
  public education. We must be vigilant. We must
• 1:48
  be responsible consumers of educational research
• 1:52
  and information. and baloney damages our field.
• 2:00
  So let's take a look at this article written
• 2:03
  by Claude Goldenberg, and there was a little
• 2:06
  bit of baloney here, part two. Goldenberg referenced
• 2:11
  two research studies to make the case for a particular
• 2:15
  type of reading instruction, and you can see
• 2:18
  the two citations there, the two studies. Now,
• 2:23
  Goldenberg cited these studies to support 21
• 2:28
  propositions. Saying it so doesn't make it so.
• 2:33
  Again, scholars do not get a free pass. They
• 2:36
  should not get a free pass on obfuscation and
• 2:40
  deception. We must call them out just because
• 2:43
  they use big words and pretty words. They do
• 2:47
  not get a free pass on these tricks. to create
• 2:53
  conveniently inaccurate perceptions of things.
• 2:56
  Now this is what Goldenberg said. As with Vaughan's
• 3:00
  intervention, Iris included explicit and systematic
• 3:04
  instruction of targeted knowledge and skills
• 3:06
  and daily lessons that include activities focused
• 3:10
  on the five core skills of phonological awareness,
• 3:13
  phonics, fluency, vocabulary, and comprehension.
• 3:16
  The intervention also included significant attention
• 3:20
  to developing oral language and integrating oral
• 3:23
  language and integrating oral language and vocabulary
• 3:26
  knowledge into the reading process. Oh my gosh,
• 3:31
  what does that mean? Sounds important doesn't
• 3:35
  it? Sounds very important. Pretty pretty words.
• 3:40
  Explicit and systematic instruction. Targeted
• 3:44
  knowledge and skills. Five core skills. Or pretty
• 3:49
  words, but it's all conveniently inaccurate.
• 3:57
  He went on to say both interventions included
• 3:59
  principles of early reading instruction consistent
• 4:03
  with findings from research on the science of
• 4:06
  reading. And again, what are these? What are
• 4:10
  the principles of early reading instruction?
• 4:12
  Who declared them to be principles? What research
• 4:16
  on the science of reading? You can't just say
• 4:20
  stuff, Claude. In describing the results, he
• 4:24
  said, effects were moderate to large on measures
• 4:27
  of English phonological awareness, word reading,
• 4:29
  decoding, reading and comprehension, text. Both
• 4:33
  interventions included principles of early reading
• 4:35
  instruction. consistent with research from the
• 4:39
  science of reading both incorporated oral English
• 4:41
  language designed as blah blah blah blah blah
• 4:44
  blah blah what the hell so let's deconstruct
• 4:51
  those sentences there was 270 words there used
• 4:56
  to create seven sentences, and there were 21
• 5:00
  propositions found within these seven sentences,
• 5:04
  and if you deconstruct them, one is led to believe
• 5:09
  that these two studies supported these 21 propositions.
• 5:17
  There was research -based evidence to support
• 5:20
  these seven things, but that is untrue. These
• 5:29
  two studies provided research -based evidence
• 5:32
  to support daily lessons that included activities
• 5:35
  focused on those five things. But that is false.
• 5:42
  And I'll be showing you that. These two studies
• 5:45
  provided research -based evidence to support
• 5:48
  giving significant attention to those four things.
• 5:52
  But if you look at the two studies, you see that
• 5:55
  is wrong. In the field of reading instruction,
• 6:01
  these are principles. And if you look, that is
• 6:06
  incorrect. It's just not true. The two studies
• 6:11
  cited by Goldenberg supported none of those propositions.
• 6:18
  Some of the words they used... included, you
• 6:22
  know, some of the descriptors used in the studies
• 6:25
  included words that were included in the proposition,
• 6:28
  but often these words were inaccurate, incomplete,
• 6:31
  or illusory, and I have a full... analysis of
• 6:36
  these two studies somewhere else, but none of
• 6:40
  the words used to indicate the 21 propositions
• 6:43
  were variables that were isolated and compared
• 6:47
  in these two studies. And that's important because
• 6:51
  he's giving the illusion, the convenient inaccuracy.
• 6:57
  the convenient, inaccurate perception that research
• 7:01
  supports these 21 propositions. Now, why is isolating
• 7:07
  variables important? In control, the experimental
• 7:11
  research and the science of reading says that's
• 7:14
  the only kind that can be used to make decisions.
• 7:18
  To determine, that's used to determine if one
• 7:22
  thing caused another thing to occur, the variable.
• 7:26
  There's a causal factor. You can say that one
• 7:29
  thing caused the other to occur. To do this,
• 7:32
  you start with two similar groups, and they're
• 7:35
  exposed to equal conditions. The treatment or
• 7:40
  the variable varies. That's the only difference
• 7:43
  between the treatment and control groups. And
• 7:46
  at the end, if there's a difference between the
• 7:49
  groups on some measure, you can say with a certain
• 7:52
  bit of confidence that the treatment was the
• 7:56
  cause of the difference. That's why it is important.
• 8:01
  Now, here's an example. Let's say I had a cage
• 8:07
  color experiment. My research question was, does
• 8:12
  the color of the mouse cage cause mice to gain
• 8:17
  weight? That's my research question. I'm looking
• 8:21
  for the causal factor. Now, in this research,
• 8:25
  I had a blue cage. But I gave all the mice marshmallows
• 8:30
  to eat there, and I had a yellow cage. And I
• 8:33
  gave the mice celery to eat in that cage. But
• 8:37
  I found out that the mice from the blue cage
• 8:40
  were much heavier after five weeks. Thus, blue
• 8:46
  cages causes weight gain, and yellow cages causes
• 8:51
  weight loss. If you wanted to lose weight, paint
• 8:54
  your house yellow. And you see the fallacy of
• 8:58
  that argument that the color of the cage was
• 9:01
  not the variable. This research did not control
• 9:05
  for what the mice ate. It was actually what the
• 9:09
  mice ate. Now, in these two studies, they varied
• 9:14
  the type of metaphorical food given to their
• 9:17
  metaphorical mice and claimed it was the color
• 9:21
  of the cage that caused improved test scores.
• 9:25
  Let me explain. Both studies compared students
• 9:28
  who received individual tutoring to students
• 9:31
  who didn't receive individual tutoring. Small
• 9:36
  group, whole class. Hmm. So one is left with
• 9:41
  three questions, none of which can be answered
• 9:44
  by controlled experimental research. Did Goldenberg
• 9:47
  not read the studies that he cited? Did he read
• 9:51
  the study to not see the confounding variables?
• 9:55
  Or did he read the studies, see the anomaly,
• 9:58
  and make the citations anyway? We don't know.
• 10:04
  In these research, the conditions given to the
• 10:08
  treatment groups were much more favorable than
• 10:11
  those given to the control groups. Tutoring,
• 10:15
  specialized training for the tutors, observations
• 10:19
  and feedback given to the tutors, professional
• 10:22
  development. All these things were in the tutoring
• 10:27
  group or in the experimental group, but not in
• 10:30
  the remediation. Thus, you cannot say it's the
• 10:34
  type of instruction. It's all these other factors
• 10:38
  that are causing the change. It is most likely
• 10:41
  that the conditions, not the type of instruction,
• 10:45
  cause the differences in scores. And any type
• 10:49
  of instruction used with the experimental conditions
• 10:52
  in these two studies would likely outperform
• 10:55
  any type of instruction used with the control
• 10:59
  conditions. Experimental group? control group.
• 11:03
  Was it the type of instruction or was it the
• 11:06
  environmental conditions? So the results. Did
• 11:09
  these studies in any way support explicit and
• 11:13
  systematic instruction? These weren't variables
• 11:16
  that were isolated. We don't know if or to what
• 11:19
  degree instruction in any group was more explicit
• 11:23
  and systematic than the other. We don't know
• 11:27
  about the type of instruction in the control
• 11:30
  groups. Did these studies support instruction
• 11:35
  of targeted knowledge and skills? And again,
• 11:40
  no. These were not variables isolated in the
• 11:44
  study. They were simply descriptors. We don't
• 11:47
  know exactly what knowledge and skills were targeted,
• 11:50
  and we don't know if the control group was any
• 11:54
  more targeted or less targeted than another.
• 11:58
  Did these studies in any way support instruction
• 12:02
  that focused on core skills? No. These weren't
• 12:06
  variables isolated. We don't know if or what
• 12:10
  degree the comparison groups contain these things.
• 12:14
  The descriptions found in the article were not
• 12:17
  aligned with research related to teaching specific
• 12:20
  core skills. We just don't know. Did both interventions
• 12:26
  include principles of early reading instruction
• 12:28
  consistent with findings from the research on
• 12:31
  the science of reading? These weren't identified.
• 12:35
  They weren't variables isolated. We don't know
• 12:38
  what the findings are. We don't know how consistent
• 12:41
  these principles are. And we don't know to what
• 12:43
  degree the groups differed in their consistency.
• 12:47
  So why did Goldenberg write these things in his
• 12:51
  article? don't know. Conclusion. Both these studies
• 12:58
  meet the basic science of reading research standards
• 13:01
  I've described, yet both are examples of bad
• 13:05
  research expertly conducted. And this bad research
• 13:11
  was creatively interpreted by Goldenberg to support
• 13:15
  the illusion of the efficacy of skills -based
• 13:18
  reading instruction for ELs. And I only analyzed
• 13:22
  one small portion of this article simply because
• 13:26
  I was familiar with the two studies. And imagine
• 13:29
  if I would go through the whole article, what
• 13:32
  I would find. In our scholarly writing, there's
• 13:37
  a big difference between using words to describe
• 13:40
  things in a study or in describing a study in
• 13:45
  isolating variables in a research study. None
• 13:49
  of the words Goldenberg used were variables isolated
• 13:53
  in these studies. None of those words can be
• 13:55
  supported by research. Neither study supported
• 13:59
  the contention that improved reading scores were
• 14:02
  caused by explicit and systematic instruction
• 14:06
  of targeted knowledge and skills and daily lessons
• 14:09
  that included activities focused on the five.
• 14:12
  core skills. This just isn't true. At best, we
• 14:18
  can say that Goldenberg used these words with
• 14:21
  extreme in precision. We must critically evaluate
• 14:28
  all scholarly work. Don't be snookered by long
• 14:32
  sentences and important sounding words. Pull
• 14:35
  them apart to see what the authors are actually
• 14:38
  saying or not saying. Examine the propositions.
• 14:42
  Examine the citations. This has been The Reading
• 14:46
  Instruction Show. I am your host, Dr. Andy Johnson.