Monday, June 16, 2014

Does homeschooling inform accurate education reform modelling?

The other week I tried to do some armchair speculating about which standard science based evolution of religion model would best fit education.  Besides having some idle fun, the goal was to investigate feasibility of an approach to analyzing the physics giving rise to education's (moralistic and ideologically influenced) reform resistant dynamics.


Last week I used blended learning as a foil for analysis.  This armchair analysis suggested that, in post-secondary contexts, education fit a performance modelled religion (or quasi-religious like group) better than a commitment modelled religion (or quasi-religious like group); A sharing of common experience seems to matter more than whether students avoid "taking the easy road out" through perceptionally easier online alternatives.

Extending this method of analysis into K-12 is problematic (for broad-based methodologies). In K-12 contexts, blended learning courses and programs tend to be extremely hybridized (see Cuban's 2013 discussion on Ed's tendency for hybridization & his blog).  Few K-12 schools clearly fence off the delivery of online courses.  Those that do, tend to either be fully online schools, or engage in much more hybridization than they are likely to admit.  The Christensen's institute's Rise of K-12 Blended Learning report (and database) and the Blend My Learning database provide lots of extremely useful information about the blended models K-12 schools employ.  Schools which offer true blended programming (fully online courses which are distinct from face-to-face or blended courses) certainly could be used to analyze the religious models best fitting education.  However, homeschooling offers another potential foil for analysis.

Ways Homeschooling Can Function as an Investigative Foil

Mixed Programming

The first way homeschooling can act as investigative foil is with regards to mixed programming between a traditional school and the family.

A small, but growing percentage of homeschooling programs (particularly fully funded, curricularly based "distance learning" programs) have agreements with parent school districts and local schools that enable "homeschool" students to take one or more face-to-face courses at their local traditional brick and mortar school.  At least in my neck of the woods, bureaucratic rules preclude this for non-curricular based homeschool programs.

A comparison of teacher and community perceptions of mixed homeschooling and pure homeschooling  has the potential to tease out distinctions between commitment model fit and performance model fit.  This could be done using methodology similar to that outlined in the blended learning analysis.  Basically, ask people what they think about homeschoolers who take all, most, some or none of their courses in a traditional school.  Measure perceived levels of acceptability versus percentage of courses done in a traditional way.

Mixed Programming Conclusion

My (unfounded) suspicion, based on a handful of experience with such cases, is that one would see a partial linear relation emerge for mixed homeschooling programs.  I also suspect a single traditional school course would make a huge difference in "acceptability" (see left graph).

This is almost what I theorized might happen in post-secondary.  In post-secondary I theorized, after one face-to-face course, there would be little difference in perceived acceptability. In other words, either things would be much less linear during the transition phase or the slope would be much less significant (see images to the left).

This may suggest K-12 schooling may have more influence from a commitment model and its hypothesized (though certainly not verified) linear effects. However, this is a pretty big "may" (in fact a pretty big double "may").

Commitment models of religion need not be any more or less linear than a performance model of religion.  Both are likely subject to phase change once a minimum input threshold has been met.  Unfortunately, I haven't seen any research on curve types.  Hence the utter speculation currently involved in such (testable) inquiries.

One thing to note from these curves is that near total acceptability in K-12 is likely only to come when all courses are completed in a traditional format.  In post-secondary, I doubt very much whether anyone cares if you took some of your electives or even core courses in an online format.  That is very much not the case with homeschooling.

Experience suggests rejection of traditional school based power structures elicit nasty community judgments.  Homeschooling stigma seems to hold whether or not families reject just the structure of schooling or the structure and curriculum.  I'd also note that this stigma seems to vanish dramatically when homeschool parents acknowledge the prime positional role of traditional schooling.  From my experience in this field, parents suffer minimal stigmatization when they indicate to the community that their reasons for homeschooling are to provide specific special needs attention not available or feasible within traditional schools.  This (untested) trend seems fits the power dynamics one expects to see with moralistic based groups (ie. religion, rule of law, strong moral ideologies, etc.).

Straight-Up Rejection Perceptions

The other way to use homeschooling as a foil for the specific science of religion model fitting education is via direct survey-styled questions.

Create a couple of different case studies that highlight various levels of homeschoolers' rejection of 1) commitment to education, 2) intent to share common (educational) experiences, 3) governmental (curricular) power.  Stitka's (2005) work suggests measuring moral convictions is possible.  A nice paper by Ryan suggests this instrument has been repeated enough to be reasonably valid.  I tend to be a bit doubtful that direct questions of moral convictions could yield the information necessary to assess the religious evolutionary model which best fits education.  However, I could be wrong.


Armchair theorizing exploring the physics associated with education's reform resistance suggests education can be interpreted through the lens of quasi-relgious dynamics.  Both blended learning and homeschooling are foils that can be used to analyze the potentially non-rational heuristics people use when judging the acceptability of others' non-traditional schooling choices.  My biased armchair analysis suggest non-linear effects may be revealed.  Non-linearity raises doubts about ration-choice decision model accuracy in educational reform contexts.

My personal familiarity with perceptions of homeschooling suggests perceived commitment levels are more of a factor in k-12 education than shared experiences.  However, the science of religion field doesn't have data about the linearity or non-linearity (curve type) of each model as a function of participant degree.  While one can assume commitment is a more linear function than performance, this is a big assumption.

Power dynamics don't fully delineate religious model fit.  Homeschooling's rejection of traditional education could either be a rejection of the power of educational ceremony (common experience) or a rejection of education's commitment power (give in to the 'borg').  My experience and my interpretation of homeschool literature (Olsen, 2008, Green & Hoever-Dempsey, 2007) suggest most homeschoolers reject education's norming (borg) role. A possible caveat is that education's norming role could be loosely interpreted as a 'norming' ceremony.  However, this really stretches the 'one-time' connotations associated with ceremony.

Therefore my conclusion is that a commitment religious model is the best fit for k-12 education while, from last week, a performance religious model is the best fit for post-secondary education.


It is interesting to see that homeschooling stigma levels are significantly greater than distance learning stigma levels despite similar degree of exclusion from the brick and mortar school houses.  It is certainly tempting to interpret this difference in terms of power dynamic.  It is also tempting to further interpret these power dynamics in terms of proximate (group) causes from multi-level theory.  And to me at least, it is hard not to contextualize the multi-level proximate causes in terms of quasi-sacred or lightly transcendental quasi-religious-like dynamics. 


Cuban, L. (2013).  Inside the Black Box of Classroom Practice:  Change Without Reform in American Education:  Harvard Educational Publishing Group

Cuban, L., Tyack. (1995).  Tinkering Toward Utopia: Harvard University Press.

Green, C., Hoover-Dempsey. (2007)  Why do Parents Homeschool?  A Systematic Examination of Parental Involvement.  Education and Urban Society, 264-285, 39 (264).

Olsen, N. (2008).  Understanding Parental Motivation to Home School: A Qualitative Case Study.  Unpublished doctoral dissertation, University of Missoula, Missoula Montana.

Ryan, T. (2013).  No Compromise: Political Consequences of Moralized Attitudes.  Paper presented at the 2013 American Political Science Association Meeting.  Chicago, Illinois.

Skitka, L. J., Bauman, C. W., & Sargis, E. G. (2005). Moral conviction: Another contributor to attitude strength or something more? Journal of Personality and Social Psychology, 88, 895–917.

Friday, June 13, 2014


Sometimes I'm asked why I would try to look at educational reform or the organizational physics/dynamics of large groups from a trajectory highly influenced by religious like dynamics?  

My answer is because it is informative,  novel, poorly understood, and real science shouldn't worry about taboos.  Another, perhaps, better framing is because moralized attitudes make a dramatic difference in the physics producing different classes of dynamics.

As Ryan (2013) says in a excellent article on the political consequences of moralized attitudes
A distinctive characteristic of moralized attitudes is that, pitted against other attitudes, they resist processing through a cost/benefit framework (Baron and Spranca 1997; Bennis, Medin and Bartels 2010; Fiske and Tetlock 1997; Tetlock 2000; Tetlock et al. 2000).

Rational approaches to educational reform dramatically miss the mark.  Educational research is, unfortunately, optimized for producing overly-precise research valid for limited, time sensitive contexts.   Teachers generally aren't known for scientific rigor.  Therefore, its important to investigate the underlying physics which give rise to the structures practitioner ed researchers describe.  Without well grasped physics, the resultant structures are wise but ephemeral happenstance.

Thursday, June 5, 2014

Education, what religious model are you?

I usually enjoy silly, online personality tests -provided of course they have just the right amount of silliness and counter-factual propositions in them.  So while I like "what famous movie are you" (Platoon & Raiders of the Lost Ark), I hate utterly loathe the pseudo science of "coloured thinking hats" and the equally pedantic "what are your animal roots". To the latter, my "… everyone else" answer flew over my wife's head...


So while browsing my amazon wish list to get a book order off, I started previewing Rossano's Supernatural Selection.  While he was rattling off the standard evolution of religion models, an interesting thought popped into my head; Where amongst the standard (evolution based) science of religion models would large morally bound & institutionalized groups fit?  More specifically, where would the institution of education fit?


Rossano categorization follows a pretty standard pattern which smartly eschews the trite adaptive vs. by-product debate.  Models are categorized as follows:

  1. commitment - individual benefit from groups which is sustained by free-loader detection routines
  2. cognitive - (over)-application of mental modules such as agent detection
  3. ecological - helps people manage the natural world (environmental regulation, existential placating, etc.)
  4. performance - ceremony and ritual bind people
  5. experiential - (not in the google preview, so I'll assume)… satiates proximate needs so it makes you feel good


My sense is people would generally apply the lens of education as a societal coherer (Tyack & Cuban 1995) resulting in a performance based religion model fit for education

A commitment model might fit education, but as I've wondered before, to what extent do weak free-loader punishment tendencies minimize fit?  It is possible to argue for a sub-set of the commitment model based on a world religion styled format: a universalist approach which accepts converts and focusses more on aquiescence than free-loader expulsion.  In that case, free loader detection may be more a case of distinguishing between true believers and luke warm marginalists: you know who the go-to people are and whose voice carries weight.


One way to theorize fit is to speculate about the predictions each model would generate with regard to free-loaders and heretics.  Light shadow systems like online education and homeschooling make convenient test points for heretical response evaluation: both delivery methods fit education's rational goal, but still elicit a bit of the disgust response common to things perceived in sacred contexts as "other".

Sitting at the edge of my desk I'd guess both online education and homeschooling present a greater challenge to education should it adhere closer to a performance based model than a commitment model.   Based upon my experience in both fields here's how I see complaints congealing:
  • Commitment model - These people get the same degree as we do but without all the work.  Taking the easy rode out is likely to eventually cause problems for everyone.  Rigor and the value of intangibles may be useful ex-post facto rationalizing arguments.
  • Performance model - They don't get the same things out of their experience as we do.  Therefore their education probably isn't as valid as mine because we really can't trust what went on.  Rigor and the value of intangibles may be useful ex-post facto rationalizing arguments.
Direct teasing of these differences is the realm of research not armchair blogging.  To me, likely rationalizations look to be practically indistinguishable without the help of sophisticated testing methods. 


The success of blended programming (online + face-to-face) in post-secondaries may illuminate things a little.  Currently blended programs are generally seen as adequately rigorous compared to completely face-to-face programs.  However, completely online programs aren't (note: a fair degree of brand association effects occur in post-secondaries that offer mixed program deliveries).  

In terms of a commitment model, this may mean a little perceived free-loading is acceptable, but there's probably a point where it crosses the line.  In terms of a performance model, it may mean blended program students have had enough common ceremony for adequacy while fully online ones haven't. 

The only difference I can see in this line of thinking is that a performance model may be accepting of blended students with just a couple of face-to-face classes while a commitment model may need a greater number of face-to-face classes.  

As a tangent, it would be really interesting to see if there is a sudden phase change in perceptions of equivalence as online course to face-to-face course ratio varies…  This would likely mean non-rational factors are at work, as pure rationality would suggest (but certainly not guarantee) a linear relation.


Thinking this over with some friends (and based upon the ed tech research) I'd wager for a performance model for education.  Students who do a couple of classes face-to-face are generally accepted as having "done their time".  Similarly, students who take a fair proportion of a face-to-face program online generally aren't interpreted as free-loading the system.

The interesting thing is that while the armchair "data" suggests a performance model for education, the rationalizations I usually see people making against online education or homeschooling seem more inline with my interpretation of a commitment model: the easy rode out will eventually cause problems for everyone.  An easy explanation for this is anyone breaking religious / group norms is seen as a troublemaker.  Norenzayan's Big Gods work on atheists is good here.


The scientific study of religion and religious like group dynamics seems to provide interesting perspectives on education's reform dynamics.  At the very least, the dynamics spinning up religious groupings seem to apply to large groups with strong cultures with moral imperatives - like education. 

Wednesday, June 4, 2014

History of Educational Reform Readings

Next year, a few teachers from a neighbouring school are doing a study session on the history of educational change.  Because almost all local PD initiatives tend to be utilitarian, I figure a change back into a bit of theory would be fun.  I'm hoping to poach some of the conversation to broach fun topics like macro-time-scale dynamic measures (Dooley & Lichtenstein, 2008) and other sundries.

With that in mind I thought I'd take a second and jot down my ideas on a intro to the topic of Ed reform historical dynamics.

Historical Philosophical Contexts
  -list, at a superficial level, some classic philosophical issues surrounding humanistic & and a-humanistic views of free will to determine the utility of coercion
- judge if the application of unstable equilibria & strange attractors explain the tensions of classical educational philosophy conundrums 

  • Hobbes vs. Russeau debate or any similar clash (this table may be a decent starter)
  • Process philosophy - say Rescher's intro to the issues (note: I'm just getting into this topic and can't say for certain how relevant this will turn out to be)
  • I'll add David Sloan Wilson's multi-level selection stuff here because it 1) captures the idea of conundrum's arising from competing tensions, and 2) it doesn't fit in any of my other categories. I'd also say that you can't fully grasp large-group dynamics without multi-level selection theory's answer to the macro-micro sociological problem.   Darwin's Cathedral is my pick for a good read here.  His New Scientist Instant Expert is a good, brief intro, but it misses the practical punch of Darwin's Cathedral. By confronting the formation of non-rational based large groups, Darwin's Cathedral helps you avoid overly utopian organization change or systemic change traps.  Thus it really helps understanding the dynamics occurring in education reform and reform resistance. 

History of modern North American educational change
-analyze historical trends in North American education reform

Complexity theory
-conceptually understand phase change, unstable equilibria and strange attractors

  • Something simple like Ball's Critical Mass or even Gladwell's Tipping Point (never read it!) to get the idea of phase change.  
  • Dooley & Lichtenstein's (2008) methodology paper on complex leadership dynamic analysis
  • There's a million pop books out on complexity, including many specific to education complexity. Many are suitable, and, surprisingly, I actually don't have any particular preferences here…

Historical dynamics
-evaluate the degree to which historical dynamics are "complex"

  • Turchin's Historical Dynamics: An intro to agent based modelling of complex situations which, at least to me, ironically shows why agent based modelling is better at inspiring creativity and forcing articulation than it is for testing and determining causal factors.
  • John Ralston Saul's Voltaire's Bastards.  This captures a bit of the evolutionary cycling down into technocracy heroic solutions.  Excellent theorizing, and it ties in nicely with Bounded Rationality.
  • While not a history book, Willis' complexity ladder gives a nice way to view cycling, especially in organizations.  While his complexity ladder is pretty much hand-waving, his conceptualization processes are very informative.
  • The Stirring of Soul in the Workplace:  A misleading title as the gist of the book is that there are always skeletons in the closet of any change.  He just expounds case study in a personal way that highlights people's good intentions. Unfortunately his other work has largely gone down Margaret Wheatley's new-age pseudo science path.
  • I think you may also need something on the free-loader problem and how that produces cyclical dynamics.  While Why we Lie got me interested in this topic, I haven't stayed current on free-loader research.  Darwin's Cathedral's view of self-interest actions as 'proximate' rather than 'ultimate' causes tends to suffice for my work. 

Bounded rationality
-comprehend that despite the possibility of seemingly high precision, accuracy is limited, especially for forward perdition

  • Haidt's Emotional dog and its rational tail paper
  • Kauffman's Reinventing the sacred (a bit too difficult for many people to grasp without prejudice)
  • Some of the Center for Inquiry's podcasts on debunking pseudo-science
  • I think I'd actually avoid some of the classic texts on traditional Bounded Rationality and Fast & Frugal heuristics.  These tackle the decision process, and to me, educational change dynamics is less about the process used to come to reform decisions and more about the causal factors producing cyclical changes and hybridized change.

Quasi-sacred or lightly transcendental cultures/organizations
-judge the degree to which non-rational factors influence change resistant groups

To me this last one is the key for deeply understanding the dynamics occurring in educational reform and reform resistance.

Case studies for analysis


Dooley & Litchenstein, 2008, pp.5
In a fractal time ecology (Koehler, 2001) the assumption is that dynamical patterns at one level of interaction are linked to emergent patterns at other levels.   Micro-scale interactions can be studied with Real-Time Observation techniques (Dooley et al., 2003);   Meso-scale interactions across days and weeks can be studied using Social Network Analysis (Wasserman & Faust, 1994), and   Macro-scale interactions across weeks, months and even longer can be studied using Event History Analysis (Poole, Van de Ven, Dooley, & Holmes, 2000)