Book Review: How Learning Works

As promised, I re­view and point-by-point sum­ma­rize How Learn­ing Works: 7 Re­search-Based Prin­ci­ples for Smart Teach­ing by Su­san A. Am­brose, Michael W. Bridges, Michele DiPietro, Mar­sha C. Lovett, and Marie K. Nor­man (2010), here­after HLW as I scratch in fu­til­ity at the sprawl­ing length of this post.

Review

The au­thors aim to provide “a bridge be­tween re­search and prac­tice” for teach­ing and learn­ing, very much in the spirit of Prac­ti­cal Ad­vice Backed by Deep The­o­ries. They con­cen­trate on widely-sup­ported re­sults that are in­de­pen­dent of sub­ject mat­ter and en­vi­ron­ment, so while the dis­cus­sion is di­rected to­wards in­struc­tors in K-12 and col­lege class­rooms, there are also im­pli­ca­tions for es­sen­tially any­one in a teach­ing or learn­ing role.

Let me restate that a lit­tle more strongly: any stu­dent, au­to­di­dact or not, would be well-served by in­ter­nal­iz­ing the mod­els and recom­men­da­tions pre­sented here. Teach­ers have even less of an ex­cuse not to read the book, which is writ­ten very clearly and with­out sink­ing to punchy pop­u­lariza­tion. This is ba­sic stuff, in the best pos­si­ble way.

Sure, there are more so­phis­ti­cated ideas out there; there ex­ist sub­gen­res of do­main-spe­cific re­search (es­pe­cially for math and physics ed­u­ca­tion); you can find di­verse per­spec­tives in home­school­ing com­mu­ni­ties or in philos­o­phy of ed­u­ca­tion. There’s even some con­tro­versy in the depths of the re­search on some of the points in this book (though for the most part the scope of dis­agree­ments is still con­tained within the bound­aries drawn by the au­thors). But as far as most peo­ple need con­cern them­selves, HLW is an earnest and ac­cu­rate if not quite com­pre­hen­sive ac­count of What We Know about learn­ing.

[I do wish there were a similar ac­count of And How We Think We Know It, look­ing into com­mon re­search tech­niques, met­rics of learn­ing out­comes, sys­tem­atic er­rors to guard against, re­li­a­bil­ity of lon­gi­tu­di­nal stud­ies, statis­tics about repli­ca­bil­ity and re­trac­tions, and so on, but this isn’t it. The book lightly de­scribes meth­ods when it sees fit, and my scat­tered checks of un­fa­mil­iar stud­ies leave me fairly con­fi­dent that the re­search does in fact bear the claims the book makes.]

The book or­ga­nizes re­search on teach­ing and learn­ing into seven prin­ci­ples in or­der to “provide in­struc­tors with an un­der­stand­ing of stu­dent learn­ing that can help them (a) see why cer­tain teach­ing ap­proaches are or are not sup­port­ing stu­dents ’ learn­ing, (b) gen­er­ate or re­fine teach­ing ap­proaches and strate­gies that more effec­tively foster stu­dent learn­ing in spe­cific con­texts, and (c) trans­fer and ap­ply these prin­ci­ples to new courses.” The prin­ci­ples are

  1. Stu­dents’ prior knowl­edge can help or hin­der learn­ing.

  2. How stu­dents or­ga­nize knowl­edge in­fluences how they learn and how they ap­ply what they know.

  3. Stu­dents’ mo­ti­va­tion de­ter­mines, di­rects, and sus­tains what they do to learn.

  4. To de­velop mas­tery, stu­dents must ac­quire com­po­nent skills, prac­tice in­te­grat­ing them, and know when to ap­ply what they have learned.

  5. Goal-di­rected prac­tice cou­pled with tar­geted feed­back en­hances the qual­ity of stu­dents’ learn­ing.

  6. Stu­dents’ cur­rent level of de­vel­op­ment in­ter­acts with the so­cial, emo­tional, and in­tel­lec­tual cli­mate of the course to im­pact learn­ing.

  7. To be­come self-di­rected learn­ers, stu­dents must learn to mon­i­tor and ad­just their ap­proaches to learn­ing.

Hope­fully these ideas are not sur­pris­ing to you. They are not meant to be; they stand mostly to or­ga­nize di­verse re­search find­ings into a co­her­ent model (see prin­ci­ple #2). And if many of those re­search find­ings are old news to you as well, I also take that to be a point in fa­vor of the book, and I trust that you will un­der­stand why.

Each chap­ter be­gins with two sto­ries meant to illus­trate the prin­ci­ple, a dis­cus­sion of the prin­ci­ple it­self, a dis­cus­sion of the re­search re­lated to that prin­ci­ple, and recom­men­da­tions that take the prin­ci­ple into ac­count. The chap­ters are in­ter­con­nected but stand on their own. If you don’t plan to teach, you might get most of your value from Chap­ters 4, 5, and 7. There’s some fluff to the book, but not much. My sum­mary, though long, leaves out the sto­ries and ex­am­ples, use­ful rep­e­ti­tions and rephras­ings, de­tailed ex­pla­na­tions, and spe­cific recom­men­da­tions, not to men­tion de­scrip­tions and cita­tions of the rele­vant stud­ies. I do not con­sider it a sub­sti­tute for read­ing the book, which isn’t re­ally that long to be­gin with.

Be­fore I sum­ma­rize HLW, I’ll make a cou­ple brief com­par­i­sons. Why Don’t Stu­dents Like School: A Cog­ni­tive Scien­tist An­swers Ques­tions About How the Mind Works and What It Means for the Class­room by Daniel T. Willing­ham (2009) looks pretty similar, down to the for­mat in which chap­ter ti­tles ask ques­tions which are then an­swered by Prin­ci­ples of Learn­ing, fol­lowed by a dis­cus­sion of the prin­ci­ple, fol­lowed by recom­men­da­tions for the class­room. It’s writ­ten at a more pop­u­lar level, with less dis­cus­sion of ac­tual re­search and lots more fluff. Only oc­ca­sion­ally does it draw con­nec­tions di­rectly to a study, rather than use that as the chief mode of ex­po­si­tion (as in HLW). Each chap­ter does have a short an­no­tated bibliog­ra­phy di­vided into less and more tech­ni­cal texts, which is nice. Willing­ham comes down strongly in fa­vor of drilling and fac­tual knowl­edge pre­ced­ing skill. While that’s some­thing I’ve ap­prov­ingly polemi­cized about at some length, it needs a moun­tain of caveats. In gen­eral he op­ti­mizes (ex­plic­itly, in fact) for coun­ter­in­tu­itive punch­i­ness, and it’s not always clear how well-sup­ported his ad­vice re­ally is. The or­ga­ni­za­tion and cov­er­age feels hap­haz­ard to me, but where he hits on top­ics cov­ered by HLW, he seems to agree.

The 25 Prin­ci­ples of Learn­ing [pdf] from the Univer­sity of Mem­phis learn­ing group is a short doc­u­ment with a similar aim: a few sen­tences de­scribing each prin­ci­ple, a cou­ple sen­tences de­scribing the im­pli­ca­tions, and a cou­ple of refer­ences. It cov­ers im­por­tant points that HLW ad­dresses only in­di­rectly or that it in­ex­pli­ca­bly leaves out en­tirely (spaced rep­e­ti­tion, test­ing, and gen­er­a­tion effects, for ex­am­ple). It’s worth look­ing over to fill in those gaps. But it’s re­ally “25 Im­por­tant Find­ings on Learn­ing”: it doesn’t give ex­am­ples, offer very spe­cific ad­vice, or at­tempt to or­ga­nize these prin­ci­ples into a causal model of learn­ing. Con­sider them ex­er­cises for the reader.

Summary

1. How Does Stu­dents’ Prior Knowl­edge Affect Their Learn­ing?

Stu­dents link new ideas and in­for­ma­tion to what they already know. This can hin­der learn­ing in the case of in­ac­tive, in­suffi­cient, in­ap­pro­pri­ate, or in­ac­cu­rate knowl­edge, but it can also be har­nessed to en­hance learn­ing.

Re­search con­sen­sus:

  • In some ways this is com­mon sense—for ex­am­ple, in the way a math­e­mat­ics lec­ture di­rectly re­lies on defi­ni­tions and the­o­rems. A stu­dent with­out suffi­cient back­ground knowl­edge might still learn to ma­nipu­late the sym­bols, but with more effort, worse re­ten­tion, worse trans­fer, and worse abil­ity to ex­plain. But there are also in­di­rect, non-ob­vi­ous mechanisms at work, in which back­ground knowl­edge that is not ex­plic­itly pre­req­ui­site can help learn­ing (as in gen­eral knowl­edge of soc­cer en­hanc­ing re­call of ar­bi­trary soc­cer-match scores).

  • Declar­a­tive knowl­edge (ob­ject-level con­cepts) and pro­ce­du­ral knowl­edge (how and when to ap­ply those con­cepts) do not always go hand in hand. One with­out the other is a knowl­edge gap that can be tricky to spot, es­pe­cially in self-as­sess­ment.

  • Ex­ist­ing knowl­edge needs to be ac­tive to be effec­tive; ac­ti­va­tion can be achieved with minor prompts and re­minders, as well as ques­tions de­signed to trig­ger re­call.

  • Stu­dents may ac­ti­vate ex­ist­ing knowl­edge that’s in­ap­pro­pri­ate (e.g. the col­lo­quial/​in­tu­itive mean­ing of “force” when learn­ing New­to­nian physics) or in­ac­cu­rate. Such ac­ti­va­tion in­terferes with learn­ing, leads to in­cor­rect con­clu­sions, and pre­dis­poses stu­dents to re­sist con­flict­ing ev­i­dence.

  • Inac­cu­rate iso­lated facts can be un­learned through em­piri­cism and ex­plicit re­fu­ta­tion. Deeper mis­con­cep­tions can be ex­tremely per­sis­tent, but pa­tient rep­e­ti­tion and a long se­ries of small in­fer­en­tial bridges can help.

Strate­gies for teach­ers:

  • Deter­mine the ex­tent, qual­ity, and na­ture (e.g. declar­a­tive vs. pro­ce­du­ral) of stu­dents’ prior knowl­edge:

    • Talk to pre­vi­ous instructors

    • Use di­ag­nos­tic tests

    • Ask self-as­sess­ment questions

    • Use brain­storm­ing or con­cept mapping

    • Look for pat­terns of error

  • Ad­dress gaps in prior knowl­edge:

    • Iden­tify for your­self what knowl­edge is necessary

    • Re­me­di­ate in­suffi­cient knowl­edge as de­ter­mined above

  • Ac­ti­vate rele­vant prior knowl­edge

    • Ex­plic­itly point out connections

    • Use analo­gies and examples

    • Use ex­er­cises that ex­plic­itly ask stu­dents to use their prior knowledge

  • Avoid ac­ti­vat­ing in­ap­pro­pri­ate prior knowl­edge:

    • High­light the bound­aries of what knowl­edge is ap­pli­ca­ble, ei­ther ex­plic­itly or with rules of thumb

    • Ex­plic­itly iden­tify dis­ci­pline-spe­cific conventions

    • Show where analo­gies break down and ex­am­ples don’t generalize

  • Help stu­dents re­vise in­ac­cu­rate knowl­edge:

    • Ask stu­dents to make and test predictions

    • Ask stu­dents to jus­tify their reasoning

    • Help stu­dents prac­tice us­ing knowl­edge meant to re­place misconceptions

    • Allow suffi­cient time

2. How Does the Way Stu­dents Or­ga­nize Knowl­edge Affect Their Learn­ing?

Devel­op­ing ex­per­tise re­quires rich con­nec­tions be­tween var­i­ous facts, con­cepts, and pro­ce­dures, or­ga­nized around ab­stract prin­ci­ples and causal re­la­tion­ships. Although an ex­pert does not nec­es­sar­ily build such knowl­edge net­works ex­plic­itly or con­sciously, it is pos­si­ble for a novice learner to de­liber­ately or­ga­nize knowl­edge into ex­pert-style struc­tures, im­prov­ing learn­ing, perfor­mance, and re­ten­tion.

Re­search:

  • The op­ti­mal or­ga­ni­za­tion of knowl­edge de­pends on how that knowl­edge is to be used. Learn­ing physics in a his­tor­i­cal frame­work has ad­van­tages and dis­ad­van­tages when com­pared with learn­ing the same physics ac­cord­ing to phys­i­cal prin­ci­ples.

  • Stu­dents whose knowl­edge net­works (graphs with “pieces of knowl­edge” as nodes linked by their re­la­tion­ships) are more densely con­nected will re­trieve their knowl­edge faster and more re­li­ably, and are more likely to no­tice in­con­sis­ten­cies and con­tra­dic­tions.

  • Ex­perts, as a re­sult of their densely con­nected knowl­edge net­works, pro­cess in­for­ma­tion in co­her­ent chunks where novices pro­cess in­di­vi­d­ual bits of in­for­ma­tion (as for chess po­si­tions and cir­cuit di­a­grams). Me­moriza­tion of digit se­quences can be greatly boosted by hi­er­ar­chi­cal chunk­ing of sub­se­quences. Th­ese facts are seen as re­lated.

  • Ex­pert knowl­edge net­works have more mean­ingful con­nec­tions and deeper or­ga­niz­ing prin­ci­ples.

  • Stu­dents learn bet­ter when pro­vided with a struc­ture for or­ga­niz­ing in­for­ma­tion. Causal re­la­tion­ships are es­pe­cially effec­tive or­ga­niz­ing prin­ci­ples.

  • Study­ing worked ex­am­ples, analo­gies, and con­trast­ing cases helps stu­dents or­ga­nize their knowl­edge mean­ingfully.

Strate­gies:

  • Or­ga­nize the ma­te­rial:

    • Create a con­cept map for the ma­te­rial to be taught

    • Iden­tify the knowl­edge or­ga­ni­za­tion best suited to the pur­pose of learning

  • En­hance stu­dents’ knowl­edge or­ga­ni­za­tion:

    • Ex­plic­itly de­scribe the or­ga­ni­za­tion of ma­te­rial at each level in the hi­er­ar­chy of pre­sen­ta­tion—sub­ject, course, lec­ture, discussion

    • Use con­trast­ing and bound­ary cases

    • Ex­plic­itly point out deep similar­i­ties and other connections

    • Use mul­ti­ple or­ga­niz­ing structures

  • Ex­pose stu­dents’ knowl­edge or­ga­ni­za­tion

    • Ask them to draw a con­cept map

    • Use a sort­ing task

    • Look for pat­terns of mistakes

3. What Fac­tors Mo­ti­vate Stu­dents to Learn?

Stu­dents are mo­ti­vated by the sub­jec­tive value of a goal and by their ex­pec­tancy of suc­cess. [You may be re­minded of the Pro­cras­ti­na­tion Equa­tion, which also de­scribes penalties for im­pul­sive­ness and de­lay.] Stu­dents may be guided by differ­ent goals, and rec­og­niz­ing this can help you foster their mo­ti­va­tion.

Re­search:

  • Stu­dents who pur­sue learn­ing goals, which em­pha­size the in­trin­sic or in­stru­men­tal value of ma­te­rial, are gen­er­ally the most mo­ti­vated and have the best learn­ing out­comes.

  • Stu­dents may also be guided by perfor­mance goals, re­lated to their self-image and rep­u­ta­tion. Th­ese may them­selves be perfor­mance-ap­proach or perfor­mance-avoidant; the former seems to en­tail a cog­ni­tive frame­work more con­ducive to learn­ing.

  • Work-avoidant goals (“do as lit­tle work as pos­si­ble”) can be di­rectly at odds with learn­ing, but are gen­er­ally con­text de­pen­dent.

  • There are, broadly, three broad de­ter­mi­nants of sub­jec­tive value: at­tain­ment value (satis­fac­tion from mas­tery or ac­com­plish­ment), in­trin­sic value, and in­stru­men­tal value. Th­ese may mu­tu­ally re­in­force each other.

  • To be mo­ti­vated, a stu­dent should ex­pect both their own abil­ity to suc­ceed and for suc­cess to bring about a de­sired out­come.

  • Ex­pec­tancy of suc­cess is in­fluenced by the stu­dent’s past suc­cess rate in similar situ­a­tions, and even more strongly by the rea­sons the stu­dent iden­ti­fies for their past suc­cess or failure. Speci­fi­cally, at­tribut­ing suc­cess to in­ter­nal and con­trol­lable causes* and failure to con­trol­lable but tem­po­rary causes in­creases ex­pec­tancy. At­tribut­ing suc­cess to luck and failure to per­sonal in­ad­e­quacy de­creases ex­pec­tancy. [*In­ter­est­ingly, the au­thors make no real dis­tinc­tion here be­tween in­ter­nal and con­trol­lable causes for suc­cess, which is a fun­da­men­tal dis­tinc­tion be­tween the “fixed” vs. “malle­able” (which you may know as “growth”) mind­sets ad­dressed in Chap­ter 7.]

  • Sup­port­ive en­vi­ron­ments also in­crease mo­ti­va­tion.

Strate­gies:

  • Estab­lish value:

    • Con­nect ma­te­rial to stu­dents’ interests

    • Provide au­then­tic tasks

    • Show rele­vance to stu­dents’ aca­demic lives

    • Show rele­vance of gen­er­al­iz­able skills

    • Iden­tify and re­ward what you (as the in­struc­tor) value

    • Ra­di­ate enthusiasm

    • Give stu­dents op­por­tu­ni­ties to re­flect on the value of their work

  • Build ex­pec­tancy:

    • Clar­ify the course goals and your in­struc­tion and as­sess­ment strategies

    • Iden­tify and set an ap­pro­pri­ate level of challenge

    • Help stu­dents build suc­cess spirals with early challenges

    • Provide feed­back on progress

    • Be fair

    • Help stu­dents at­tribute suc­cess and failure appropriately

    • Dis­cuss effec­tive study strategies

  • Give stu­dents flex­i­bil­ity and con­trol in course work to in­crease both value and expectancy

4. How Do Stu­dents Develop Mastery?

Con­sider a driver chang­ing lanes, mak­ing many small mo­tions, vi­sual checks, and men­tal eval­u­a­tions fluently and au­to­mat­i­cally. An ex­pert performs com­plex tasks with lit­tle con­scious aware­ness of the com­plex­ity in­volved. To ap­proach that level of mas­tery, a novice must not only learn the com­po­nent skills, but also in­te­grate the skills and know when to ap­ply them.

Re­search:

  • Ex­perts do not nec­es­sar­ily make good teach­ers: they pro­cess in­for­ma­tion in chunks, they em­ploy short­cuts and skip steps, they perform with au­to­mat­ic­ity, and they over­es­ti­mate stu­dents’ com­pe­tence. Their un­con­scious mas­tery leads to so-called ex­pert blind spots.

  • Stu­dents will perform poorly if their com­po­nent skills are weak.

  • Stu­dent perfor­mance is greatly im­proved when in­struc­tors iden­tify com­po­nent skills re­quired for a com­plex task and tar­get weak ones through prac­tice. A small amount of fo­cused prac­tice on a com­po­nent skill can have a large im­pact on perfor­mance of the com­plex task.

  • Mul­ti­task­ing de­grades perfor­mance by way of ex­cess in­for­ma­tion-pro­cess­ing de­mands or cog­ni­tive load. The same ap­plies to com­bin­ing skills for a com­plex task, but much more so for novices than for ex­perts.

  • Cog­ni­tive load can be re­duced when learn­ing a com­plex task by al­low­ing the stu­dent to fo­cus on one com­po­nent skill at a time. It may also be helpful for the in­struc­tor to sup­port other as­pects of the task while stu­dents do their fo­cused prac­tice. This is known as scaf­fold­ing.

  • Another in­stance of scaf­fold­ing effect ap­pears when the in­struc­tor pre­sents stu­dents with worked ex­am­ples rather than prob­lems, free­ing up cog­ni­tive re­sources to think about prin­ci­ples and tech­niques.

  • Re­sults on drilling com­po­nent skills in iso­la­tion, as com­pared with prac­tic­ing the over­all task with fo­cus on the com­po­nents, are mixed. Some skills af­ford iso­lated prac­tice bet­ter than oth­ers. A highly com­plex but eas­ily di­visi­ble task can be learned more effec­tively by ini­tially prac­tic­ing the com­po­nents in iso­la­tion, and then pro­gres­sively com­bin­ing them.

  • Mastery also in­volves know­ing when to ap­ply learned skills out­side of the learn­ing con­text. Do­ing so is referred to as trans­fer. Trans­fer oc­curs rarely and with difficulty, and is worse the more dis­similar the learn­ing and trans­fer con­texts.

  • Over­speci­fic­ity and con­text-de­pen­dence of knowl­edge hurt trans­fer; deep un­der­stand­ing of prin­ci­ples and re­la­tion­ships helps trans­fer. The lat­ter effect can be tar­geted with struc­tured com­par­i­sons and analog­i­cal rea­son­ing also help trans­fer.

  • Minor prompts and re­minders fa­cil­i­tate trans­fer, much as they help ac­ti­vate ap­pro­pri­ate knowl­edge (see Chap­ter 1).

Strate­gies:

  • Ex­pose com­po­nent skills:

    • Map out your own ex­pert blind spot

    • En­list help from those with mere con­scious competence

    • Talk to oth­ers in your discipline

    • Talk to oth­ers out­side your discipline

    • Ex­plore ed­u­ca­tional materials

  • Re­in­force com­po­nent skills

    • Fo­cus stu­dents’ at­ten­tion on the key as­pects of the task

    • Di­ag­nose weak or miss­ing com­po­nent skills

    • Provide iso­lated prac­tice of those skills.

  • Build fluency and fa­cil­i­tate in­te­gra­tion of skills

    • Give stu­dents prac­tice ex­er­cises ex­plic­itly to in­crease automaticity

    • Tem­porar­ily con­strain the scope of the task

    • Ex­plic­itly in­clude in­te­gra­tion in perfor­mance criteria

  • Fa­cil­i­tate trans­fer:

    • Dis­cuss con­di­tions of applicability

    • Give ex­er­cises ex­plic­itly about con­di­tions of applicability

    • Provide op­por­tu­ni­ties to prac­tice in di­verse contexts

    • Use hy­po­thet­i­cal sce­nar­ios for prac­tice questions

    • Ask stu­dents to gen­er­al­ize to ab­stract principles

    • Iden­tify deep fea­tures us­ing comparisons

    • Prompt stu­dents to re­trieve rele­vant knowledge

5. What Kinds of Prac­tice and Feed­back En­hance Learn­ing?

Prac­tice is of­ten mis­guided and feed­back poorly timed, in­suffi­cient, or un­fo­cused. To be effec­tive, prac­tice should be di­rected by goals and cou­pled with tar­geted feed­back.

Re­search:

  • Learn­ing can be pre­dicted by time in de­liber­ate prac­tice, which is marked by be­ing di­rected to­ward a spe­cific goal and an ap­pro­pri­ate level of challenge. [I’ve of­ten heard de­liber­ate prac­tice de­scribed with an em­pha­sis on mind­ful at­ten­tion, in con­trast with prac­tice in a flow state (for ex­am­ple in an ar­ti­cle by Eric­s­son him­self—the last para­graph be­fore “Fu­ture Direc­tions”), but the au­thors ques­tion­ably sug­gest that flow is a sign of ap­pro­pri­ate challenge. For mo­ti­va­tion, per­haps it is, but I would ar­gue not so for de­liber­ate prac­tice.]

  • Clearly speci­fied perfor­mance crite­ria can help di­rect stu­dents’ prac­tice.

  • Learn­ing is ham­pered by ei­ther in­suffi­cient or ex­ces­sive challenge.

  • The suc­cess of in­di­vi­d­ual tu­tor­ing is largely driven by the abil­ity to tai­lor challenges to a level ap­pro­pri­ate to de­liber­ate prac­tice.

  • An in­struc­tor can im­prove learn­ing out­comes with difficult tasks by adding struc­ture and sup­port to bring it within the bounds of the stu­dent’s com­pe­tence. This can con­sist of guidance by the in­struc­tor, or of writ­ten prompts and check­lists. (C.f. “scaf­fold­ing” in Chap­ter 4.)

  • The benefits of de­liber­ate prac­tice ac­crue grad­u­ally with in­creas­ing time spent prac­tic­ing; both stu­dents and teach­ers un­der­es­ti­mate the time needed.

  • The effec­tive­ness of feed­back is de­ter­mined by both con­tent and timing. It should com­mu­ni­cate progress and di­rect sub­se­quent effort, and it should be sup­plied when stu­dents can best use it.

  • Feed­back that iden­ti­fies spe­cific items that need im­prove­ment will aid learn­ing more than will a mere in­di­ca­tion of er­ror.

  • Un­fo­cused feed­back can be coun­ter­pro­duc­tive by over­whelming the stu­dent and failing to di­rect effort well.

  • Gen­er­ally, more fre­quent and more rapid feed­back is bet­ter for learn­ing. De­layed feed­back can be use­ful in helping stu­dents learn to rec­og­nize and cor­rect their own er­rors.

Strate­gies:

  • Estab­lish goals:

    • Be ex­plicit about course goals, and phrase them in terms of ca­pa­bil­ities rather than knowledge

    • Use a rubric to com­mu­ni­cate perfor­mance criteria

    • Give con­trast­ing ex­am­ples of high and low qual­ity work

    • Pro­gres­sively re­fine goals

  • En­courage de­liber­ate prac­tice:

    • Assess prior knowl­edge to set an ap­pro­pri­ate challenge

    • Create many chances to practice

    • Build scaf­fold­ing into assignments

    • Set ex­pec­ta­tions about practice

  • Tar­get feed­back:

    • Look for pat­terns of errors

    • Use pri­ori­tized feed­back to di­rect stu­dent efforts

    • Give feed­back on strengths and weaknesses

    • Allow fre­quent op­por­tu­ni­ties for feedback

    • Provide feed­back at the group level, po­ten­tially in real-time

    • Re­quire peer feed­back on assignments

    • Re­quire stu­dents to de­scribe how they in­cor­po­rated feedback

6. Why Do Stu­dent Devel­op­ment and Course Cli­mate Mat­ter for Stu­dent Learn­ing?

Peo­ple vary not just in­tel­lec­tu­ally, but also so­cially and emo­tion­ally. Stu­dents’ iden­tities may be en­tan­gled with the course ma­te­rial and en­vi­ron­ment in com­pli­cated ways that of­ten go un­rec­og­nized. A stu­dent’s en­tire state—not just the in­tel­lect—in­ter­acts with the so­cial, emo­tional, and in­tel­lec­tual cli­mate of the course to im­pact learn­ing, for bet­ter or for worse. [When I saw this chap­ter ti­tle, I had a vague worry that it would seem out of place, a per­func­tory nod to di­ver­sity stud­ies or some­thing. I’m still not en­tirely com­fortable with parts of the treat­ment here, but the above premise is sound.]

Re­search:

  • The re­search in­volved in this first sec­tion is of a differ­ent na­ture from the rest of the text. In the first part, the au­thors seek to de­scribe stu­dent de­vel­op­ment, and cite a model which char­ac­ter­izes de­vel­op­men­tal changes into seven di­men­sions: de­vel­op­ing com­pe­tence, man­ag­ing emo­tions, de­vel­op­ing au­ton­omy, es­tab­lish­ing iden­tity, free­ing in­ter­per­sonal re­la­tion­ships, de­vel­op­ing pur­pose, and de­vel­op­ing iden­tity. They then cite re­search char­ac­ter­iz­ing in­tel­lec­tual de­vel­op­ments in terms of stages: du­al­ity, mul­ti­plic­ity, rel­a­tivism, and com­mit­ment. Similarly, stages for so­cial de­vel­op­ment. The point is that peo­ple can have a lot of differ­ent im­plicit and ex­plicit be­liefs, modes of com­mu­ni­ca­tion, and ways of pro­cess­ing new in­for­ma­tion, which they can’t just switch off and ho­mog­e­nize when they en­ter a class­room, and that peo­ple have done a lot of work to at­tempt to enu­mer­ate and con­nect these things. [I think the dis­cus­sion here is the weak­est part of the book, and I’d be in­ter­ested in bet­ter re­sources on the sub­ject, if they ex­ist.]

  • For course cli­mate, they de­scribe a clas­sifi­ca­tion in terms of whether an en­vi­ron­ment is marginal­iz­ing or cen­tral­iz­ing (de­scribing how the per­spec­tives of groups might be dis­cour­aged or wel­comed), and whether this oc­curs im­plic­itly or ex­plic­itly. Im­plic­itly marginal­iz­ing class­rooms are the most com­mon of the four quad­rants.

  • In im­plic­itly marginal­iz­ing en­vi­ron­ments (i.e. with­out overt ex­clu­sion or hos­tility to­wards out­groups), in­di­vi­d­u­als may suffer an ac­cu­mu­la­tion of micro-in­equities that over time has a large im­pact on learn­ing. A num­ber of stud­ies have found that per­cep­tions of a marginal­iz­ing cli­mate are nega­tively cor­re­lated with learn­ing and ca­reer out­comes. The au­thors iden­tify four im­por­tant chan­nels for marginal­iza­tion: stereo­types, tone, fac­ulty-stu­dent in­ter­ac­tions, and con­tent.

  • The ac­ti­va­tion (in the sense of Chap­ter 1) of stereo­types can in­fluence learn­ing, gen­er­ally im­pairing perfor­mance; this effect is known as stereo­type threat. The ac­ti­va­tion does not have to be a re­sult of ex­plic­itly in­vok­ing the stereo­type; im­plicit com­mu­ni­ca­tion of as­sump­tions or ap­par­ently in­nocu­ous com­ments also have effects.

  • The im­me­di­ate mechanism for stereo­type threat seems to be a dis­rup­tive emo­tional re­ac­tion; this as op­posed to self-effi­cacy or self-es­teem be­ing de­pressed or oth­er­wise brought in line with the stereo­type. The effect does not re­quire any be­lief in the stereo­type. There are deeper nu­ances as well as strate­gies for miti­gat­ing the effect in the liter­a­ture.

  • Per­ceived hos­tility or ex­pec­ta­tions of failure in stereo­types can de­crease mo­ti­va­tion and drive stu­dents from a dis­ci­pline.

  • A pos­i­tive, con­struc­tive, and en­courag­ing tone in dis­cus­sions and syl­labi im­proves stu­dent mo­ti­va­tion and be­hav­ior. (This in con­trast to puni­tive, crit­i­cal, or de­mean­ing tone.)

  • Per­ceived pos­i­tive fac­ulty at­ti­tudes to­wards and in­ter­ac­tions with un­der­grads are cor­re­lated with higher rates of grad­u­ate ed­u­ca­tion and bet­ter self-re­ported learn­ing out­comes. Fac­ulty availa­bil­ity is a ma­jor fac­tor in stu­dents’ aca­demic de­ci­sions.

  • Course con­tent it­self in its ori­en­ta­tion to­wards in­clu­sive­ness can have cog­ni­tive, mo­ti­va­tional, and so­cio-emo­tional effects on learn­ing.

Strate­gies:

  • Pro­mote in­tel­lec­tual de­vel­op­ment:

    • Make un­cer­tainty, am­bi­guity, and com­plex­ity safe

    • Re­sist a sin­gle right answer

    • In­cor­po­rate use of ev­i­dence into perfor­mance criteria

  • Pro­mote so­cial de­vel­op­ment:

    • Ex­am­ine your as­sump­tions about your students

    • Be mind­ful of ac­ci­den­tal cues re­gard­ing stereotypes

    • Do not ask in­di­vi­d­u­als to speak for an en­tire group

    • Rec­og­nize stu­dents as in­di­vi­d­u­als.

  • Pro­mote an in­clu­sive cli­mate:

    • Be a model for in­clu­sive lan­guage, be­hav­ior, and attitudes

    • Use mul­ti­ple and di­verse examples

    • Estab­lish and re­in­force ground rules for interaction

    • Make sure course con­tent does not marginal­ize students

    • Use the syl­labus and first day of class to es­tab­lish climate

    • Set up pro­cesses to get feed­back on the climate

    • An­ti­ci­pate and pre­pare for sen­si­tive issues

    • Ad­dress ten­sions early

    • Turn dis­cord and ten­sion into a learn­ing opportunity

    • Fa­cil­i­tate and model ac­tive listen­ing.

7. How Do Stu­dents Be­come Self-Directed Learn­ers?

As one pro­gresses in aca­demic and pro­fes­sional life, one takes pro­gres­sively more re­spon­si­bil­ity for one’s own learn­ing. The jump be­tween high school and col­lege can be es­pe­cially jar­ring in this re­gard. Me­tacog­ni­tion, “the pro­cess of re­flect­ing on and di­rect­ing one’s own think­ing,” be­comes in­creas­ingly im­por­tant, but falls out­side the scope of most in­struc­tion. Still, to effec­tively di­rect their own learn­ing, stu­dents must learn and prac­tice an ar­ray of metacog­ni­tive skills.

Re­search:

  • One model rep­re­sents metacog­ni­tion as a con­tin­u­ously loop­ing cy­cle of task as­sess­ment, eval­u­a­tion of strengths and weak­nesses, plan­ning, ex­e­cu­tion and si­mul­ta­neous mon­i­tor­ing, and re­flec­tion; all of these five steps are in­formed by a stu­dent’s be­liefs about in­tel­li­gence and learn­ing.

  • Assess­ing the task is not always nat­u­ral or ob­vi­ous to stu­dents (es­say prompts are of­ten ig­nored; learn­ing goals are not always clear).

  • Peo­ple are poor judges of their own knowl­edge and skills, tend­ing to over­es­ti­mate their abil­ities more the weaker they are.

  • Novices spend lit­tle time in the plan­ning phase of the cy­cle rel­a­tive to ex­perts in physics, math, and writ­ing. Novice plans are of­ten poorly matched to the task.

  • Stu­dents who nat­u­rally and con­tin­u­ously mon­i­tor their perfor­mance and un­der­stand­ing learn bet­ter.

  • Stu­dents can be taught to self-mon­i­tor, and this also im­proves learn­ing.

  • Mon­i­tor­ing alone is not suffi­cient; novice prob­lem solvers will con­tinue to use a strat­egy af­ter it has failed (and cer­tainly af­ter it has proven mod­estly suc­cess­ful and fa­mil­iar but not op­ti­mal).

  • Stu­dents who be­lieve their in­tel­li­gence is malle­able rather than fixed are more likely to learn and perform well.

  • More­over, the “malle­able” per­spec­tive can be pro­moted by ex­ter­nal in­fluences, still lead­ing to bet­ter perfor­mance.

Strate­gies:

  • Pro­mote task as­sess­ment:

    • Be more ex­plicit about as­sign­ments than you think is necessary

    • Tell stu­dents what you do not want

    • Check stu­dents’ un­der­stand­ing of the task in their own words

    • Provide a rubric

  • Pro­mote self-eval­u­a­tion:

    • Give timely feedback

    • Provide op­por­tu­ni­ties for self-as­sess­ment.

  • Pro­mote plan­ning:

    • Have stu­dents im­ple­ment a plan you provide

    • Have stu­dents im­ple­ment their own plan

    • Make plan­ning the cen­tral goal of the as­sign­ment.

  • Pro­mote self-mon­i­tor­ing:

    • Provide sim­ple heuris­tic ques­tions for self-evaluation

    • Have stu­dents do guided self-assessments

    • Re­quire stu­dents to re­flect on and an­no­tate their own work

    • Use peer review

  • Pro­mote re­flec­tion and ad­just­ment:

    • Prompt stu­dents to re­flect on their performance

    • Prompt stu­dents to an­a­lyze effec­tive­ness of study skills

    • Pre­sent mul­ti­ple strategies

    • Create as­sign­ments that fo­cus on strategizing

  • Pro­mote use­ful be­liefs about in­tel­li­gence and learn­ing:

    • Ad­dress these be­liefs directly

    • Broaden stu­dents’ un­der­stand­ing of learning

    • Help stu­dents set re­al­is­tic expectations

  • Pro­mote metacog­ni­tion:

    • Model your metacog­ni­tive pro­cess for your students

    • Scaf­fold stu­dents in their metacog­ni­tive processes

Con­clu­sion: Ap­ply­ing the Seven Prin­ci­ples to Ourselves

The au­thors turn their prin­ci­ples in­ward and dis­cuss learn­ing to teach. For the most part this is a restate­ment of the prin­ci­ples with no par­tic­u­larly new in­sights in their ap­pli­ca­tion to teach­ing, but there are in­ter­est­ing com­ments re­gard­ing the first few:

  • Many teach­ers were formerly atyp­i­cally suc­cess­ful stu­dents, and their prior knowl­edge can lead to dis­torted ex­pec­ta­tions; ac­cord­ingly, many of the recom­men­da­tions in­volve gath­er­ing data about the stu­dents.

  • The or­ga­ni­za­tion of this book into prin­ci­ples is it­self a de­liber­ate ap­pli­ca­tion of the sec­ond prin­ci­ple.

  • For mo­ti­va­tion, the au­thors try to con­nect the con­tent of the course with what ev­ery teacher re­ally cares about: effi­ciency. They also sug­gest fo­cus­ing on one or two as­pects of teach­ing in a given semester, in or­der to build up small suc­cesses in im­prov­ing teach­ing.

  • In terms of mas­tery, prac­tice and feed­back, cli­mate, and metacog­ni­tion, teach­ing is not so differ­ent from any other skill.

Appendices

HLW has eight ap­pen­dices on tools men­tioned through­out the book, with a re­it­er­a­tion of their na­ture and util­ity, and most im­por­tantly, ex­am­ple check­lists and work­sheets. Th­ese are

  • Stu­dent self-assessment

  • Con­cept maps

  • Rubrics

  • Learn­ing objectives

  • Ground rules [for dis­cus­sion]

  • Exam wrap­pers [for pro­mot­ing metacog­ni­tion on graded ex­ams]

  • Checklists

  • Reader re­sponse/​peer review

Th­ese alone would have been an im­prove­ment over most teach­ing ma­te­ri­als I grew up with.