1978 on interpreting the effects of repetition - solving a problem versus remembering a solution

8/9/2019 1978 on Interpreting the Effects of Repetition - Solving a Problem Versus Remembering a Solution

1/19

JOURNAL

OF

VERBAL

LEARNING

AND

VERBAL

BEHAVIOR

7, 649_667

(

1978)

On Interpreting

he

Effects

of

Repetition:

Solving

Problem

Versus

Remembering

Solution

Lnnny

L. Jecosy

M ac

M uster

Ll niuersitt,

When

a

problem

is repeated.

the later presentation

of

the

problem

sometimes

results

in the

subject esponding

by remembenng

he solution

rather

han by

going

through.the

operations

ha t

would

otherwise

be

necessaryo solve

he

problem.

The means

of

obtaining the

solution

is shown

to influence

subsequent retention

performance:

retention

of the

solution suffers

if it

has been

obtained

by remembenng

ather

than

by solving

the

problem.

The

distinction

between

olving

a

problem

and remembenng

a

solution is

used n

an account

of the

etTect f

spacing repetitions

and

other

standard

memory

phenomena.

The relevance

of the

distinction

to tasks

such

as

word

perception

is also discussed.

Suppose

hat

you

are

asked o

find

the sum

of

37

+

15+

12.Af te rhav ingobta inedrh issum

your

are

mmediatelypresented

ith

the

same

problem.

The type

of

processing

hat

you

do

will

differ

drastically

on the repeated

pre-

sentation.

On the first

encounter

you

un -

doubtedly

went

throu_eh

the

process

of

addit ion

to

obtain

the

sum:

on the

second

encounter.hesum s readilyavailable nd ca n

be

_uiven

without

going

back

through

th e

operations

of adding

the numbers.

ndeed.

a

full

repetit ion

of the

processing

ctivity

ma y

be

diff icult,

f

not impossible,

o

accomplish

without

some

delay. which

is

probably

th e

rationale

or

the commonly prescribed

outine

of

checking

an

addition

by

adding

the

num-

bers in

reverse

order

rather

than

simply

re-

adding

them in

the

same

order.

To

make it

possible

o repeat

he full

process

f addit ion,

it is

probably

sufficient

o

separate

he

re -

The

author

is

grateful

o

Lee Brooks

and

F. l .

M. Craik

for

comments

and suggestions

n an earl ier

draft

oi this

paper.

This

researchwas

supported

by

Grant 4028

t from

the

National

Research

Council

ol Canada.

Address

repnnt

requests

to

Larrv

L.

Jacoby.

Deparrment

oi

Psychology.

McMaster

Universi ty.

Hami l ton.

Onrario.

Canada

L8S 4Kl .

649

petit ions

of the

problem

by

severalntervening

problems

of the

same orm.

This

example

of

addit ion

is

the

basis

of the

analysis

of

the

eflect

of reperit ion

on

memory

that is

presented

n

this

paper.

The

task

of

memorizing

a ist

of

words

can

becompared

o

the task

of

solving

a series

of

problems

Th e

presentation

f a

word

[or

memory

constitutes

a problem: the subjectmust find operarions

that wil l

render

that word

memorable

after

some

delay.

For

example.

the

subject mav

image

the referent

of the

word

in

order

to

enhance

memorv.

As

with

math

problems.

t is

unlikely

hat

a repetit ion

of a

word

results

n

a

full

repetit ion

of

the

processing.

f

one haslust

imaged

their

own

dog in

order

to

make

the

word

dog

more memorable.

magin_e

heir

dog a

second

ime

as

a consequence

f

th e

word

being

repeated

s

unlikely

to require

a

iull

repetit ion

of the

processes

hat

were

necessary

or

the

original maging.

n

general.

it seems hat one can retrieve he product of

their

prior

memorizing

activiry

without

fully

repeatin,e

hat

memorizing

activiry.

The

means

by

which

a solution

o

a

problem

is

obtainedwill

influence

ubsequent

erention

of the

problem

and i ts

solut ion.

This

cla im

has been

used

n recommending

discovery

.oor,,*n, ltt,

i'

f \31 ]ll?)];.il' i

l l r ights

o i rcproductron n any. ' brm

rescrred.


2/19

650

LARRY L. JACOBY

learning

as compared

o

reception

earning

(Bruner,

1966);

he

suggestion

s

that working

through

a

problem

to its

solution enhances

memory

ascompared o a situationwhere he

solution is

provided.

Litt le is

known

about

how

these

effects

work.

However,

one

possi-

bil ity

is

that

solving

a

problem

results

n

a

richer

memory

of

that

problem

and its

solution.

In

the math

example,

the

further

operations

hat

are

performed

when

addition

is required

may

result

in

a

more

extensive

memory

of

the

problem

by

including

substeps

leading

o its

solution.

This

additional

infor-

mation

could

provide

a further

basis for

subsequent

ecognition

of the

problem

an d

increase

he

number

of

potential

cues or

later

recall A secondexplanationof the retention

advantage

of solving

a

problem

as

compared

to reading

or effortlessly

remembering

the

solution

appeals

o the

role

of consciousness

in

determining

subsequent

etention.

n

th e

math example.

addin-e

series

f numbers

o

obtain

a solution

invloves

consciousness

n

a

way

that

effortless

rememberin-e

of the

solution

does

not. When

adding the

numbers.

it

seems

necessary

o

monitor

one's

ow n

processing

while

an effortless

etrieval

of a

solution

seems

automatic .

The

involve-

ment

of

consciousness

may

enhance

sub-

sequent etentionperformance.

This

analysis

s

relevant

o

the

spacing

ffect

that is wel l

documented

n

the

memory

itera-

ture

(Hintzman.

19741.

he

ar-eument

s

that

the

proccssing

of the

first

presentation

of a

word

makes

avai lable

an

appropriate

en-

coding

and

thereby

rivializes

he

processing

associatedwith

the

second

presentation

of

the

word.

As

the

spacing

of repetit ions

s

increased.

he

amount

of

processing

of

the

repeatedword

that

is required

to

attain

an

appropriate

encod

ng

increases:

onseq

ent y,

one should

expect

etention

o

be

enhanced s

a function

of

the

spacing

of

repetit ions.

As

ar-euedwith

reference

o solving

a

problem.

working

with

an item

to

derive

an encoding

produces

subsequent

etention

that

exceeds

that

produced

when

an appropriate

encoding

is

effortlessly

etrieved.

The experiments

hat

are to

be reported

provide

a clear

demonstration

oi the

memorv

consequences f solvin_s problem versus

remember ing

solut ion.

Much

of the

subse-

quent

discussionwill

center

around

the

effects

of spacing

epetit ions.

However.

he

contrast

between

olving

a

problem

and

remembering

a solut ion

is

appl icable

over

a much wider

range

of

si tuat ions

han is

usual lv

onsidered

when

discussing

he

mernor iz ing

f

a l ist

of

words.

One

potent ial

appl icat ion

har

is

of

current

interest

nvolvesword

identitrcation.

A

pronunciation

for

a

word

can be

con-

structed

by

goin_e

hrou_eh

series

I rules

ha t

deal

with

letter

o

sound

correspondences.

s

in the math example however. th is con-

structive

activity

is

likely

to be

bvpassed

r

minimizedwhen

the

conditions

are such

as o

allow

the subject

o

easily remember

a

pro-

nuniciation

that

he

has

encountered pre-

viously.

Thus,

he

contrast with

which

we

ar e

dealing s

elevant

o many

tasks n

addition

to

those

of solving

math

problems

or memoriz-

ing word

lists

Potential

applications

of the

distinction

between

solving

a

problem

an d

remembering

a

solution

are

described

n

th e

-eeneral

iscussion.

The

_general

iscussion

also includes

a

review

of

severalexperiments

hat

can

be

used

to

support

the

claim

that

an advantage

n

subsequent

etention

s

gained

by

construct-

ing

rather

than

remembering

a solution.

Th e

ar*qument

hat the

effect

f

spacinu

epetit ions

resuits

rom

a change n

the

mode

of obtainins

a

solution,

or

achieving

an

encoding.

s

ex -

panded

and

contrasted with

other

explana-

tions.

This

ar_qument

s

then

extended

to

account

or

var ious

memorv

phenomcna

hat

have

previouslv

been

discusscd

n

rnuch

narrower

contexts.

The

dist i r lc t ron

bctncen

solv inga problemund rcmcnrhrcr i r lg.r olu-

t ion

is

shoun

tt l have

cr lnsidcrublc

cur ist ic

value:

his

dist inct ion

can bc

uscd

o suggest

exper iments

hat w'ould

not

arise from

the

more

radi t ional

explanat ions

f the

phenom-

ena

that

are reviewed.


3/19

RETENTION

EFFECTS

OF

SOLVING

VS

REME}IBERINC

6 5 1

ExprnryrNr

I

Subjects n_ea_eed

n

a task hat is

similar o

that of solvinga crosswordpuzzle. cueword

was

presented

lon_e

ith

a few letters

and

a

series

of blanks

representing

he

missing

letters

of a

word

that was

related

o the cue

word

(e.g..

oot

s-

-e).

The

subject's ask

wa s

to

report

the

word

that

could

be

produced

by

f i l l ing

he blanks

shoe

n

the

aboveexample) .

In

some

instances.

he

task

of solv ing

the

puzzle

was

trivialized

by

preceding

he

prob-

lem

with

i ts

solut ion: he

pr imary

manipu-

lation in

the

first

experimentwas

to vary

th e

spacin_u

f the

puzzle

and its

solution.

Th e

processing

equired

o

obtain the

solut ion

and, consequently,

ater

memory were

ex -

pected

o

be

_greater

hen presentation

f the

solution was

separated rom

the

puzzle

by

intervening

items

rather

than immediately

preceding

he

puzzle

in

the

list. Retention

performance

was

assessed

y means

of an

unexpected

est

of cued recall:

he

cue

word

from

each

of

the

puzzles

(e._s.,

oot) was

presented

s a

cue

for

recall

of the

solution

words.

Compar isons

among

condi t ionswere

de-

signed

o

provide

nformation

about

the

pro-

cessing arried out to solve the puzzle.Fo r

example,

n

the first

experiment.

ued

recall

af ter eading

he

solut ion

and hen

solving

he

puzzle

was

compared

to cued

recall

after

having

read

the

solut ion

twice.

When

the

solution word

immediately

preceded

resen-

tation

of the

puzzle,solving

he

puzzle

was

no t

expected

o

entail

any

more

effort or

produce

any better

recall

than

would

result

f lrom

simply reading

he

solution word

for

a second

t ime.

With

greater

eparat ion

f a

puzzle

nd

its

solut ion.

however.

the requi rement

of

solvin_e

he

puzzle

was

expected

o

produce

hi-eheretention han would be producedby a

second

eadine

of the

solut ion

word.

Methotls

Desig4n

ntl

suhjecrs.

ubjects ither

read

or

const ructed

he ight-hand

member

f

pai rs

of

related

words.

For

pairs

hat

were

o be read.

the r ight-hand

member o i the

pai r

was pre-

sented

intact.

For

pai rs

that requi red a

response o

be constructed, two

or more

letters

were

deleted rom the interior

of the

right-hand

member

of the

pair:

he

subject

wa s

to

say

the

word

that could

be formed

by

restoring

he missing etters.

The experimentwas

designed

o that each

of s ixcondi t ions

were

epresented

y

l l i tems

mixed n

a s ingle

72- i tem

ist .

One condit ion

(R)

consisted

of

the

12 i tems

that

were

presented

nly

onceand in

which

the response

had

only to be

reutl

by the

subject.A

second

condit ion

(C)

consisted

f the l2 i tems

that

were presented nly once but for which th e

response

ad

to be

c'orr.srrurc'tetl. ln

wo

oi

th e

remaining

onditions,

ach

pair

waspresented

twice with

the response

eing

ead

both times

(RR):

in

one

of these RR

condit ions. he

second

resentation

mmediately

ollowed

he

first,and in the

other

it

lollowed

with

a lag of

20 i tems. n

the f inal

two

condi t ions

he tem

was

to be read

the

first t ime

and constructed

the

second

RC):

again. n

one

of the RC

condi t ions

he repet i t ionwas

mmediate

nd

in

the

other

al ter 20 intervening

tems.

Eighteen

ubjectswere

paid

S2.00ihour

o

participate.Testine was conducted

in

indi-

vidual

sessions.

Muteriuls.

Seventy-two

pairs

of related

words

were

selected rom

the

Connecticut

free-association

norms.

In

selecting

pairs.

neither

he

most frequent

association

o

a cue

word

nor

a bizarre

associationwas

selected.

The intent was

to

select

pairs

such

that the

responseword

could

be so lved in

the con-

dit ions

requi r ing

construct ion without

the

solut ion

being

too

obvious.

The response

members

f

pairs

varied

n length

rom

four to

eight

etters.

When

constructionof a response

was

required.

he

first etter

and

the ast etter

of the response ere

always

presented.

or the

longer responsewords,

up to

four letters

including

he

first

and

last letter

oi the

word

were provided:

two

or more letters were

deleted rom

each esponse ord

that

required


4/19

652

construction.

Deleted

etters

were

eplaced

y

blanks

so

that

the

number

of let ters

n

the

word that was o be constructedwasobvious.

Six l ists

were

formed

by rotating pairs

through presentat ion

condi t ions

so that

across

ists

each

presentation

ondition was

represented

y the

same

pairs.

Within

a

list.

the

order

o[

pairs

was

such

that

each

pre-

sentation

ondition was

epresented

y

rrpairs

before

any

presentation

ondition was

repre-

sented

bv

rr

+

I

pairs.

Proc'etlure.

he list

of

pairs

was

prepared

as

a

stack

of

note

cards

rvith

each

note

card

containing

one re latedpair.

A

t iming

device

was

used

o

pace

ubjects

hrough

his

stack

of

notecardsat a rateof 6 seconds/card ubjects

were

nformed

hat we

were

nterested

n

ho w

long t

took

them

o

solve

roblems

f

the ype

they

might

encounter

n

a crosswordpuzzle.

They

were

o turn

a note

card

when

signaled

to do

so by

the

imin_e

evice.

f

the ri_eht-hand

member

of the

pair

on the

note

card was

no t

intact.

hey

were

o

say

a

word

that

contained

the

provided

letters

and whose

remainin,s

letters would

fit in

the

blanks:

they

were

further

informed

that

the

response

hey

,eave

had

to be related

o

the

cue

word

that was

provided

on the.card.

s

soonas hey hought

they

knew

the

answer.

hey were

to

push

a

button

that

was

n

front

of them

and

say

th e

solut ion

aloud

f

the

ri_eht-hand

ember

of

the

pair

on

a

card

was

ntact

heywere

o

push

the

button

and read

the

response

aloud.

Subjects

were

old

that

their

reaction

imes

o

read

items were

to

serve

as

a baseline

or

their

reaction

ime

to responses

hat

had

to

be

construced.

n reality.

eaction

imeswere

no t

recorded:

he reaction-time

ask was

simply

used o

provide

a

cover

story

for

subjects.

After

subjects

had worked

their way

through he deckof notecards.heywere_eiven

an unexpected

est

of cued-recall;

he

eft-hand

member

of each

pair

was provided,

written

in

a random

order

on a

sheet,

s

a cue or

recall

of

the right-hand

member

of

each

pair.

The

cued-

recall

est

was

subject-paced.

Anulvses.

he

test

of

cued-recallprovided

LARRY L. JACOBY

the

data that

are of

pr imarv

interest.

n

analvzing

the

cued-recall

data

comine from

condi t ions hat have been requi red o con-

struct

a

response.

one has

the

option

oi

conditionalizing

cued-recall

on successful

construction

of

the response.

he rationale

or

such

condit ional izin_s

s: I f

a

subject

was

unable

to

construct

a

part icu lar

response

dur ing

presentat ion

f the is t hen

he

was

not

exposed

o that

response:

onsequently.

he

subject

cannot

be expected

to recall

th e

response

on

the later

test

of cued

recall.

Despite

his

consideration.

he

probabil ity

of

cued-recall was

not

conditionalized

in

th e

analyses

hat

are to be reported

he

decision

not

to

condi t ional ize

he

probabi l i ty

of cued-

recall

was

motivated

by concern

or

a

poten-

tial

confoundin_e

hat

could result

rom

item

selection

roblems.

Cued-recall

an

obviously

not

be

conditionalized

when

subjects

only

read

the

solution

to a

problem;

conditional-

'

izing

or

the

problems

hat required

construc-

t ion

of

a solut ion

may

resul t

in

select ive ly

droppin_e-out

he harder

pairs

so that

th e

comparison

of the

read

and

construct

conditions

s

confoundedwith

the

diff iculty

of

the

pairs

on

which

their

performance

is

assessed. lthough they will not be reported.

analyses

were

also

carried-out

with

con-

dit ionalized

scores.

n

_eeneral,

he

result

of

conditionalizing

scoreswas

to

make

effects

larger

hat were

also

present

nd

signif icant

n

the

analyses

f unconditionalized

scores.

n

no

instance

did the

results

of

an analysis

of

conditionalized

scores

conflict with

con-

clusions

hat

are o

be drawn

from

an analysis

of unconditionalized

cores.

The

level

of signif icance

or

all

statistical

testswas

set

at

p


5/19

spaced) the

rate

of successfulconstruction

was

909/., ignificantly higher: F(1, 171:13.60.

MS.-.011. When the i tem was to be con-

structed immediately

after

having been

read

(Read-Construct,

immediate) the rate

of

success

was

99i(,

significantly

higher

than

the

delayed

condition: F(1,

17): 23.94,

MS.

:.gtg3.

The argument made earlier

was

that

con-

structing a

response s

a solution

to a

problem

should

produce

retention

greater

than

that

produced

by simply reading

the response.

Further, the

retention

advantage

hat

would

result from

solving a

problem

should

depend

on the

processing

nvolved in constructing he

solution. mmediatelyprecedinga problem by

presentation

of its

solution

should trivialize

the

problem

to such an extent that the

pro-

cesses

nvolved in

solving the

problem

should

not

differ appreciably from those that are

required to simply read the

solution

a second

time; consequently,

one should expect no

advantage to result from construeting the

solution

as

compared

to a second eading of

the solution.

When

presentation

f the solu-

tion is widely

separated from that of

the

problem,

however,

solving of the

problem.

should be nontrivial

and

give

rise o retention

that exceeds hat

coming

from reading

the

solution for a

second time. The cued-recall

data are

presented

n Figure I

and

provide

support for each

of the above

predictions.

When

a

pair

was presented

nly once,con-

struction

of

a

solution esulted n

substantially

higher

cued-recall

han did

simply reading he

solution

word

in

a

pair,

F(1, 171:55.92,

MS.:.02. For

pairs

that

were

repeated. he

effectof spacing epetitions

was

much

greater

in the Read-Construct

condition than in the

Read-Read

condition, F(1, 17\:22.00,

MS.: .91. When reading of the solut ion

immediately

preceded presentation

of the

problem,

cued recall in

the

Read-Construct

condition did not

exceed

hat

in

the Read-

Read

condition; however, with

spaced

presentation,

the Read-Construct condi-

tion

produced

substantially higher

cued

653

-ts

RC

85

25

IMMEOIATE

SPACED ONCE

PRESENTED

Frc.

J.

Probabil i ty

of cued-recall

as a

function

of

reading

(R)

vs

construction

(C)

in E.rperiment

L

recall than did the Read*Read condition.

Comparisons

with

once-presented

items

reveal hat reading he solution

immediately

prior

to being

required to construct

th e

solution

produces

lower

performance

than

results

rom constructing he solution

without

havin_e

reviously

read it, F(1. L7)- 15.91.

MS.: .91.

With

spaced

presentat ions.

he

retention advantage conferred by a

prior

readingof the solution

n the Read-Construct

condition

is

approximately

equal to that

gained

n the Read-Readcondition.

That is.

the

difference

between

Read and Read-Read

is approximately equal to

that

between

Construct

and Read-Construct:

the

prior

readingof the response nhances ecall n both

instances.

It

was

once

generally

believed that the

important condit ion

for

learning

was

to lead

the subject,by

whatever means. o make a

correct esponse. his belief

n the mportance

of

making

the correct

esponse asmotivated

RETENTION EFFECTS

OF SOLVING

VS

REMEMBERING

=

f,

75

I r l

E

a6s

5

555

do5

s

lt

Pss

L


6/19

654

LARRY

L. JACOBY

educational

practice.

It

is not

unusual

for

a

teacher

to

present

a

problem

along with

its

solution

and then

require

he

class

o

parrot

that

solution.Within theSkinnerian radit ion,

programmed

instruction

was

designed

to

ensure

hat

a correct

response

was

made.

An

inserted question

often

occurs

almost

im-

mediately

after

the

text

that

provides

the

answer

o

that

question;

n

addition, prompts

such

as rhyming

cues

or

a

portion

of the

etters

comprising

the

response

are

provided

to

further

ensure

hat

the

correct

response

will

be

given.

One

point

to

be

made

by

the

present

study

is that

the

processes

nvolved

in

solving

a

problem

determine

etention

of

the solution.

If the

problem

is

trivialized

by

presenting

he

solution immediately prior to the problem or

by

simpy

requiring

the

person

to read

the

solution,

retention

performance

will

suffer.

It

might

be

argued

hat

the

retention

advan-

tage

_eained

by

constructing

rather

than

reading

or

remembering

a

solution

is

due

to

differences

n

study time;

it

takes

longer

to

construct

a

solution

than

to read

one,

and

this

difference

n

effective

study

time

is

responsible

for

effects

n

subsequent

retention.

First, it

probably

did

not

take

twice

as

long to

con-

struct

a solution

as

to read

a

solution.

How-

ever,

reading

the

solution

twice

produced

substantially ower recall than

did

construct-

ing

a

solution

only

once;

recall

of once-

presented

items

that

required

construction

was

higher

than

that in

the

Read-Read

con-

dition.

Further,

arguments

about

differences

in

efectiue

study

time

are

meaningless

nless

we

have

some

dea

of

what

constitutes

effective

study,

and of

the

variations

n

processing

hat

are

responsible

or

differences

n

the

effective-

ness

of

study.

Other

data

(e.g.,

Craik

&

Tulving,

1975)

can

be

used

to

suggest

that

differences

in

time

per

se

are irrelevant

to

differences

n retention

that

are

produced

by

manipulating

orienting

tasks.

ExpenrueNr

2

The

results

of the

first

experiment

could

be

summarized

by

the

statement

hat

increas-

ing

the effort

required

to

solve

a

problem

enhances

later

retention performance.

The

second

experiment

provides

further

evidence

on the role of effort by directly varying the

difficulty

of

the

problems

themselves.

n

one

condition,

the

crossword

puzzle

problems

were

extremely

easy

o

solve. Puzzles

or

that

condition were

constructed

by deleting

a

single interior

letter from

the

solution

word

(e.g.,

check

m-ney);

the

result

for

most

pairs

was

o

make the

problem

so easy

hat

it

seems

possible

to

just

read

the

solution word.

Puzzles

for

a

second

condition were

made

more

difficult

by

deleting

two interior

letters

from

the

solution word

(e.g.,

ance

s5r).

As

can

be

seen from

the

examples,

deleting

a

second etterappears o producea substantial

increase

n

the

difficulty

of the

problem.

The

more

difficult

problems

were

expected

o

yield

higher

retention performance

The

second

experiment

also

differed

from

the

first

in that

the

effect

of

intermediate

levels

of spacing were

investigated

in

the

second

experiment.

n

the

current

memory

literature

there s

some

disagreement

whether

there

s a

dichotomous

effect

of immediate

vs

spaced

repetitions

or

a true

continuous

effect

of

spacing

repetitions;

hat

is,

some

studies

ind

differences

only

between

massed

and nonzero

levels

of spacing

while

other

studies

find

differences

among

nonzero

levels

of

spacing

(see

Hintzrnan,

1974,

for

a

review).

In

the

present

situation,

this

observation

can

be

translated

nto

a

speculation

about

the

role

of

short-term

memory

A

prior

presentation

of

the

solution

to

a

problem

might

reduce

ater

retention

only

if

that

solution

is

still in

short-

term

memory

when

the

problem

is

encoun-

tered

If

so,

one

would

expect

a

difference

between

mmediate

and

widely

spaced

epeti-

tions

but

would

not

expect

increases

in

spacing outside the range of short-term to

influence

later

retention.

The

interaction

of spacing

with problem

difficulty

is

also

of

interest.

With

massed

presentation

of

the

solution

and

problem,

the

two

levels

of

problem

difficulty

should

yield


7/19

equivalent evels

f later recall:

or

both

types

of

problems,

he

task of

provid ing

a solut ion

should be trivial. At greater evelsoi spacing.

however,

he more

diff icult

problems

should

produce

higher

etention han the

easier nes.

Method

Desigtt

tntl

subjec'rs.he second

experiment

employed he samecrossword

puzzle

ask as

did the first

experiment.

However.all repeti-

t ion

conditions in the

second experiment

involved irst reading

he responsemember

of

a

pair

and then ater

encountering hat

pair

as

a

problem

that

required

he

previously

ea d

response s a solution (the Read-Construct

arrangement

n

Experiment

1). Eight repeti-

t ion

conditions

were produced

by lactorially

combining

two levels

of

solution diff iculty

(Easy

vs

Hard)

wi th

four

levels f spacing

f

presentations

0,

10,

20, or

40

intervening

pai rs) .

n four

addit ional

condit ions.a

pair

was

presented

only once.

To

produce

these

four

conditions,

the two

levels

of

problem

dif f icul ty

were

combined

with

the

solut ion

o

the

problem

being

either read

or constructed.

Problem

diff iculty was

a

pseudovariable,

n-

serted for

purposes

of

analyses.

when

th e

solut ions

to

the

once-presented

roblems

were

read.

All

condit ions

were

represented

within-subjects.

The

subjects

were

l8

students nro l led

at

McMaster

Universi ty

who werepaid

52.00rhr

to

part icipate

n

the experiment.

iVI

teriuls

untl

procetlure.

The

materials

comprised 120

pairs

of relatedwords

selected

from the

Connecticut ree-association

orms

using

the

same criteria

as

in

Experiment 1.

Solution

words

varied

rom four

to

six etters

in length.

Easy

problems

were

produced

by

replacingone interior letter of the solution

word with

a

blank;

difrcult

problems

were

produced

by replacing

wo

interior

letters

of

the

solution

word

with

blanks.

To

construct

a

list,

10

pairs

were

assigned

o

each of the

12

conditions

described

n the

design and

subjects

section.

Since

eight

of

655

these

2

condit ions equired epet i t ions i a

pair,

a

l ist

was

200

pai rs

long.

For the

repet i t ion ondit ions. resentat ionsf a pai r

were

separated y

0, 10,

20 or

40

intervening

pairs.

Twelve

ists

were

constructedby rotat-

ing

pairs

through condit ionsso that

across

lists each

condition

was

represented

,v-

he

same

pairs:

six of these ists

were

presented

o

two

subjects

hi le

he emaining

ix

ists

w'ere

presented

o only one

subject.

A

final cued-recall est

was

constructed n

the samemanner

as described or

Experiment

l. The

procedure

was

also

identical

to that

described

or

the first

experiment.

,, lncrlr'si.s.

s in

Experiment l.

the cued-

recall data that will be reported were no t

conditionalized

on the subject

correctly

solvin_s he

corresponding crossword

puzzle

problem.

Again.

condit ional ized

ata

were

alsoanalyzed,

ut the esults

f those nalyses

do not al ter

any conclusions

rawn on the

basis

of

the

uncondit ional iz'ed

ata.

Signif icance evel for

all tests

was

set at

p

< . 05 .

Re.su/rs

ntl Di

sc

L.ssiorr

Di f ferencesn

the

probabi l i ty

of

an

un-

successful ttempt

at

solving the crossword

problems

verif ied

that

the

hard

problems

were

indeed

more

dif f icul t than

were

the

easy

problems:

he

probabi l i ty

of being

unable to solve

a

problem

in

the

once-

presented

ondit ions

was

.12 for

hard

prob-

lemsand .02 or

easy

problems.

Prior

reading

of the

solut ion faci l i tated

solv ing

of the

problems

when

reading of the

solut ion

im-

mediately

preceded

resentation

f the

prob-

lem

(0-spacing);

he

probabil ity

of bein_s

unable o solve

problem

under

hose

i rcum-

stanceswas quite low (.005I or both the easy

and the

diff icult

problems.

When

40

items

intervened

between

eading he solution

an d

presentat ion

f the

problem.

he

probabi l i ty

of

being unable

o solvea

diff icult

problem

(.04)

was

stil l ower

than

that in

the once-presented

condition

where

the solution

was

not read

RETENTION

EFFECTSOF SOLVING

VS

RE},IEMBERING


8/19

656

70

20

40

o^rcE

PRES€NTED

SPAC/IVG

Frc.

2. Probabi l i ty

of

cued-recal l

as

a

function

of

construction

difficulty

and

spacing.

prior

to

presentat ion

f

the

problem

(.12).

With

the

easy problems.

n

contrast.

the

probabi l i ty

f

being

unable

o

solve probtem

when

40

tems

ntervened

etween

eading

he

solut ion

and presentat ion

f the problem .02)

was

equal

o

that

in

the

corresponding

nce_

presented

ondit ion.

The

cued-recall

results

lrom

the

second

exper iment

are

displayed

in

Figure

2.

A

portion

of

those

esults

imply

replicate

ffects

found

in

the

first

experiment.

Among

th e

once-presented

tems,

being

required

o

con-

st ruct

a

solut ion

produced

substant ial ly

hi_eher

ued-recall

han

did

reading

he

solu-

t i on .

F (1 .

l 7 ) -80 . 76 , VS : . 02 .

W i t h i n

t he

conditions

that

required

construction.

ead-

ing

the

solut ion

mmediate ly

r ior

to

solv ing

a problem that required that solution (0 -

spacin_e)

owered

later

cued-recall

as

com-

pared

to

the

corresponding

once-presented

condi t ions

hat

solved

he

problem

without

previously

reading

the

solut ion.

F(

.

171:32.64,

MS.

: .02.

LARRY

L. JACOBY

It was

earlier

suggested

hat

the

results

of

the

first

experiment

reflect

the

influence

of

short-term

memory.

The

suggestion

was

that

prior reading of the solut ion wil l depress

later

cued-recall

only if

the

solution

resides

in

short-term

memory

after

the

problem

requir ing

that

solut ion

is

presented.

This

posi t ion

leads

ro

the

predict ion

that

in-

creases

n

spacing

beyond

he

range

of

short-

terrn

memory

should

have

no

effect

on

later

cued

recall;

that

is,

one

should

find

an

immediate

vs

spaced

ffect

but

should

ind

no

diflerences

amone

-sreater

evels

of

spacin_e.

The

results

f

the

second

xperiment

evealed

a

quite

large

effect

of

spacin_s resentations.

f (3.

5

; :

33.98,

MS.

-

02.

Cont rary

ro ex-

pectations,

however,

the

effect

oi

spacing

presentations

emained

signif icant

when

th e

O-spacing

onditions

were

dropped

from

th e

ana lys is ,

(2 ,34 \ :3 .93 ,

MS.

- .02 .

I t

does or

seem

reasonable

o

argue

that

the

effects

of

spacin_e

ithin

the

range

of lG40

intervening

items

are

due

to

differences

n

the probabil ity

of the

solution

residing

n

short-term

memory

durin_e

he

presentation

of the problem:

th e

levels

of

spacing

nvolved

are

al l

outside

of

what

is

usually

considered

o

be

the range

of

short-term

memory.

Some

factor

thit

operates ver a _ereaterange han doesshort-

term

memory

is

apparently

responsible

or

the

spacing

effect

observed

in

the present

experiments.

Although

he

more

dif f icul tproblems

were

expected

o

produce

higher

retention

than

were

he

easyproblems.

esults

rom

the

once-

presented

tems

reveal

no

effect

of

problem

diff iculty

on later

cued-recall.

owever.

effects

of

problem

diff iculty

are

observed

when

on e

examines

the

repeated

items.

Among

th e

repeated

tems.

the

more

diff icult

problems

produced

higher

cued-recall

han

did the

easy

problems, (1. 17\:9.76 fulS,: .007. Exami-

nation

of

the

data

presented

n

Figure

2

suggests

hat

problem

diff iculty

nteracts

with

the

spacing

f

presentations.

t

0-spacin_e.

he

two

levels

of

problem

diff iculty

produced

essentially

equivalent

levels

of

cued-recall

60

50

40

. 24

20

J

H

q

( \

5

=

d

fl

F<

ti

n-

o i

,

i

t*t*'

to


9/19

while

the more

difficult

problems produced

higher

performance

han did the

easy

prob-

lems

at the

greater

levels of spacing. The

interaction

of spacing

and

problem

difficulty,

however,

was

not

statistically

signif icant,

F < 1 .

A

significant nteraction

was

ound

rvhen

an

analysiswas

carried

out on the two levels

of

problem

difficulty

at

40-spacing

and the

two

once

presented

tem

conditions hat required

const ruct ion

of a response,

(1,

l7)-5.53.

MS.: .028

Examinat ion

of this interact ion

shows that

problem

difficulty had

an

effect

with

the repeated

tems

but

not with

once

presented

tems.

Further,

prior

reading

of

the

solution enhanced ecall relative to the once-

presented

tems

only

for

the

difficult

problems:

at the longest-spacing,

he level

of cued recall

produced

by

easy items is

approximately

equal o

that

produced

by the corresponding

once-presented

tems.

These

esults

can

easily

be

nterpreted

n

the

same

erms

as

was

Experiment

1. Presenting

the solution

of a

problem

prior

to the

pre-

sentation

f

that

problem

provides

he

subject

with

two

means

of

_eenerating

response: he

subject

an either remember

he

solution hat

he

was given previously

or he

can use the

information provided by the problem to con-

struct

a solution

(mixes

of

the two

means

of

-eenerating

a response

are,

of course.

also

possible).

When

a

presentation

f the solution

immediately precedes

a

presentation

of the

problem,

he

subject

almost

certainly

emem-

bers rather

than

constructs

he

solution.

an d

later retention

performance

suffers

he effect

of

spacing

of

presentations

or

both

levels

of

problem

difficulty

can

be

interpreted

as

being

due to

a correspondin_e

ncrease

n

the likeli-

hood

that

a solution

to the

problem

must

be

constructed

rather

than remembered.

When a solutionwasnot presented rior to

the

presentation

of a

problem

as

was

the case

with

once-presented

airs.

he subject

had no

option but to

construct

a

solution. t

appears

that

the

only important

factor for

later reten-

tion

was

that

construction

be required:

he

RETENTION

EFFECTS

OF SOLVING

VS

RE}TEMBERING

657

diff icultyof

the

problem

did not influence

ater

cued-recall

performance.

This lack

of an effect

of

problem

diff iculty may

simply result

from

problem

diff iculty

having

been manipulated

over

too narrow

of a range; however.

th e

manipulation

was

sufficient

o

produce

sub-

stantiallymore unsuccessful

ttempts o

solve

the

diff icult

problems

as

compared to the

easier nes.Perhaps he most

surprising esult

is the ar_se

dvantage n

cued-recall

roduced

by

an easy

construction

as compared

to

readin-e

he

solution

to

a

problem.

As shown

by

the

example

provided

earlier,

he deletion

of a single etterappears

o make

he

problems

so easy hat

one can

just

read the

solution:

however, olvingproblems hat wereeven his

easy

produced

subsequent

recall

that

was

double that

produced

by actually

readin_e

he

solution. Additional

research

s required

to

determine

whether

or not a

continuous

effect

of

problem

diff iculty

can be

obtained. f the

effects

prove

to

be dichotomous, as

is sug-

gested

y the results

f the

present

xperiment.

it

may

be

necessary

o invoke

the

concept

ol

consciousness

o

explain

the effect

of

problem

diff iculty.

To

enhance ater

retention.

t

ma y

only

be necessary

o

disrupt

the flow

of

processing

o

that some

minimal

amount

of

conscious onstruction s required.

The

effect

of

problem

difficulty found with

repeated tems

remains

to

be explained.

n

these

cases,

roblem

difficulty

may have

had

its

effect

by

influencing

the

ease

of

remember-

ing

the solution.

Even

at the longer

spacings.

subjects may

have

sometimes

remembered

rather than

constructed

the

solution.

This

rememberin_s

f the solution

is more

likely

with

the

easy

problems

where

only one

letter

of the solution word

is

deleted

han

with

th e

hard

problems

where

wo

letters

of the

solu-

tion

are

deleted.

That is.

because here

ar e

more

letters

and

therefore

a

more restrictive

context.

the

easy

problems

provide

a

better

cue or

recall

of the

previously

_eiven

olution

than

do

the

hard

problems:

consequently.

constructon

of

the

solution is

required

more

often for

the

hard

problems

with

result ing


10/19

658

better

retention.

Evidence

that

the

prior

reading

of the

solution

does

nfluence

solving

the problem even at the longest spacing is

provided

by

both

experiments

n

both

experi-

ments,

he

probabil ity

of

being

unable

o

solve

a

problem

was

lower

when

the

solution

had

been

read

previously.

This

reduction

in

the

probabil ity

of

being

unable

o solve

a

problem

is

presumably

due

to

the

solution

being

at

least

partially

remembered

n

some

nstances

rather

than

being

solely

constructed.

The

possibility

of

remembering

ather

than

constructing

a

solution,

even when

the

solu-

tion

does

not

immediately

precede

presen-

tation

of the

problem,

casts

a new

ight

on the

role of short-term memory in producing the

effect

of

spacing

epetitions.

Greeno

(1967)

has

emphasized

he

role

of short-term

memory

in

producing

the

spacing

effect

by

arguing

that

a

subject

might

learn

nothing

from

the

pre-

sentation

of

an

item

if

that

item

currently

resides

n

short-terrn

memory.

This

is

said

to

be

because

he

subject will

not

select

a

new

code

for

an item

that

resides

n

short-term

memory

during

its

repetition;

memory

over

the

long

term

is

described

as

requirin_e

he

selection

f

an

appropriate

code.

Similarly,

n

the

present

paper.

it was

suggested

that

presentation

of

a

problem

may

have

litt le

effect

on

subsequent

etention

f

the

solution

to

that

problem

currently

resides

n

short-

term

memory

so

that

solving

the

problem

is

trivialized.

On

the

basis

of the

results

of

th e

present

experiment,

however,

t

appears

hat

effortless

ememberin_e

ather

than

residence

in

short-terrn

memory

is

the important

factor

for

subsequent

retention.

Discussions

of

short-term

memory

have

usually

emphasized

limited-capacity

notions

so

that

it

is

th e

number

of intervening

tems

that

is

seen

as

determinin_e hether or not a particular item

will

still

reside

n

short-terrn

memory

when

it

is

repeated.

mplicating

ease

o[

remembering,

in

contrast,

emphasizes

he

importance

o[ the

cues

provided

for

retrieval

of

an

earlier pre-

sented

olution

as

well

as

he

number

of items

intervening

between

presentation

f

the

solu-

LARRY

L.

JACOBY

t ion

and

that

of the

problem.

An

implication

of emphasizing

retrieval

is

that when

re-

membering of the solution is enhanced by

providing

more

effective

cues.

as in

the

easy

construction

as compared

to

the

hard

con-

struction

conditions,

subsequent

retention

performance

will

suffer

even when

presen-

tation

of the

solution

does

not immediately

precede

hat

of the

problem.

The

presentation

of less

effective

ues or

retrieval

makes

t

more

likely

that

the

subjectwil l

have

to

solve

th e

problem

rather

than

remember

he

solution.

and

subsequent

etention

benefits.

The

above

account

of

the results

claims

that

remembering

a

solution

always

eads

o

poorer later remembering

of that

solution

than

does

construction

of the

solution.

Such

a

position

is

too

extreme

n that

rememberin_s

sometimes

nvolves

construction.

As

one

ex-

ample,

Lindsay

and

Norman

(197'7)

argue

convincin_ely

hat

construction

or reconstruc-

tion

is involved

when

we

answer

a

question

about

where

we

were

on

some

specified

ata

in

the

distant

past.

Perhaps

a

distinction

needs

o

be

drawn

between

effortful

and

effortless

retrieval

(e.g.,

Gotz

& Jacoby,

1974).

Effortful

retrieval

nvolves

many

of the

sameprocesses

as

doesconstructionand acts he sameway as

construction

to

enhance

ater

retention.

In

contrast,

effortless

emembering

of a

solution.

regardless

f

the

spacing

of the

solution

an d

problem,

s

much

ike

reading

he

solution

an d

does

elatively

itt le

to

enhance

ater

retention

performance

Further

theorizing

at

this

point

is

by

necessity

speculative

However,

one

advantage

offered

by

the

procedures

em-

ployed

in

the

present

experiments

s

that

the

task

is

one

that

can

be further

analyzed

to

yield

information

about

the

processes

n

which

subjects

engage

o

deal

with

a

problem.

The main questions eft unanswered n the

above

account

are: What

is

involved

in

th e

construction

of

a

solution

and

why

does

engaging

n

construction

enhance

ater

re-

tention performance?

These

questions

will

be

considered

n

the

_seneral

iscussion.

Before

considering

those questions,

however. the


11/19

RETENTION

EFFECTS

OF SOLVING

VS

RE}IEMBERINC 659

spacin_s

ffect

btained here

will

be compared

with

that obtained in

more typical memory

experiments, nd the applicabil ity of current

theories

oi the spacing

effect o the results

of

the

present

experiments

will

be discussed.

Gexrnel

DrscussroN

Whereas

t

is

possible

hat

the

spacing

ffect

found

here has

a totally

different

basis han

does he

spacing

effect ound in

more

typical

memory

experiments

e._e..

elton. 1967),

t

seems

more

likely

that

the two

are

closely

related.

n

order

o memorize word,

a subject

must

engage

n

some

series

f operations:

or

example, inding relations among words or

ima*uing

he words.

As with

themath

problems

considered

earlier

and

the

crossword

puzzle

problems

used n

the

present

experiments.

t

seems

nlikely

that these

memorizin_s

pera-

tions

are

fully

repeated when

the

second

presentation

f a

word

immediately

ollows

ts

first

presentation.

n

the

remainder

of

this

paper,

I

procecd

as if

the

spacing

effect ound

here

and

the

spacing

effect ound

in

more

typical

memory

exper iments

ave

a

common

basis. f

this

common

basis s

accepted,

t is

of

interest

o see

how var ious

theories

of the

spacing

ffect

are

n

attempting

o

account

or

the

results

of

the

present

experiments.

One

explanation

of

the spacing

effectha s

appealed

to

differences

n the

frequency

of

rehearsal

as

a function

of

the

spacing

of

repetit ions.

The

claim

is

that

an item

is

rehearsed

durin_e

he

interval

intervening

between

its

presentations;

consequently,

spaced

epetit ions

of

an item

result

n

more

rehearsal

f the

repeated

tem than

do massed

repet i t ions

Rundus.1971).

his

grearer

um-

ber

of rehearsals

s

used to

explain

th e

retentionadvanta_uef spaced epetit ionsby

assuming

hat

lon_e-term

emory

of

an tem s

a

direct function

of the

number

of

rehearsals

that i tem

has

eceived.

l though

t

may

apply

in

other

situations,

he

frequency

f

rehearsal

explanation

cannot

account

ior

the

spacing

effect

obtained

in

the

present

experimenrs.

First. the incidental

learnine

procedure

employed

here made t

uniikely

that subjects

would rehearse n item during intervalsout-

s ide f i ts

presentat ion.

ore mportant ly.

he

differential ehearsal

xplanation

cannot ac-

count for the

debil itatin-e ffect

f reading he

solution to a

problem

immediately

prior

to

solving the

problem.

[t is not

reasonable o

claim that

the

prior

reading

of the

solution

resulted

n

the

solution being

rehearsed

es s

than it

would

have been

had the

solution nor

been read

prior

to

presentation

f the

prob-

lem.

The

encoding var iabi l i ty

hypothesis

has

provided

a second

explanation

of

the

effect

f

spacing epetit ions.By this hypothesis. here

are severaldifferent ays

a to-be-remembered

word

can be encoded:

he

more

different

wavs

a

word

is

encoded

he

better

will

be

retention

since

each different

encoding

provides

an

additional

access

route

to

the

word

in

memory.

t

is

urther

assumed

hat an ncrease

in

spacin_e

akes

it

more likely

that

repeti-

t ions

o[ an item

will

be

encoded

differently.

Thus,

the effect

of

spacing

of repetit ions

s

attr ibuted

to

an increase

n

the

number

of

encodings

f

the repeated

tem

(Melton.

1961

Madigan.

1969).

A

variant

of the

encoding

variabi l i ty

hypothesis

ssumes

hat an i tem

becomes

onditioned

to contextual

elements

that are

active

during the

presentation

f

th e

item.

The

spacing

effect s

then

explained

on

the

basis

f differences

n

the

similarity

of these

contextual

elements

as

a function

of spacinu

(Anderson

&

Bower.

1972:

Glenberg,

1971).

There

seems

o be no

way

that

anything

ik e

encodin_e ariabil ity

could

have

operated

o

produce

the

spacing

effect observed

n

th e

present

xperiments.

he encoding

variabil ity

hypothesis

appears

irrelevant

when

on e

abandons he procedure f presenting list of

words

to

be

memorized

and instead

presents

a series

oi

problems

that

are to

be solved.

Notions

discussed

arlier.

however,

do

pro-

vide

a means

of reinterpret in g

ata

that have

been

presented

s

support ing

he

encodinu

variabi l t ty

explanat ion

of

the

spacing

effecr.


12/19

660

LARRY

L .

JACOBY

Several

nvest i_sators

e.g.,

Madigan.

1969)

have

demonstrated

hat

the

effect

of

spacin_u

repetit ions

can

be

reduced

by

varyin_u

he

context in which the repeated word is

presented.

f

the

context

biases

a

different

interpretation

for

each presentation

of the

repeated

word

(e.*e..

ever-CHILL.

snow-

CHILL)

the

spacing

ffect s

f lat ter

han

it is

when

he

context

b iases

he same

meaning

or

each

presentat ion.

h is

biasing

of

dif ferent

interpretat ions

by

manipulat ing

conte.\ t

s

assumed

o

mimic what

happens

n

ordinary

ci rcumstances

hen

epet i t ions

f an

item

are

widely

spaced:

t

is

c la imed

hat

both

manipu-

lat ions

ncrease

he

number

of

access

oures

o

the

repeated

tem.

An

alternat ive

nterpre-

tat ion.

however

s

that

the

change

n

context

essentially

produces

different

problems

that

are

to

be

solved.

Changing

context

s

anal -

ogous

o

f irst

asking

a

person

o

add

37and

l5

and

then

asking

hem

to

mul t ip ly

37

and

15.

Al though

he

numbers

emain

he

same

n

the

two

problems.

he

answer

o the

first

problem

cannot

be

carr ied

over

o

t r iv ia l ize

he

solvin_s

of

the

secondproblem

Simi lar ly.

operarions

carr ied

out

to

encode

an

i tem

n

one

context

may

not

provide

an

encodin_e

hat is

appro-

priate

to

the item

repeated

n

a

different

context.The manipulat ionof context resul ts

in

more

hr l l

processing

f

later

presentat ions

of the

repeated

tem.

and

consequent ly.

n-

hances

etent ion.

A

third

explanat ion

f

the

spacing

ffect

s

simi lar

to

the

account

offered

here.

By

a

hab i tua t ion

ypothes is

H in tzman.

974) .

he

spacing

effect

s

due

ro

the

deficient

regis-

trat ion

of laterpresentat ions

hen

epet i t ions

of an i tem

are

massed:

intzman.

Block.

and

Summers

1973\

provide

evidence

hat

the

encodinc

of later presentat ions

s

def icient.

This

deficient

regisrrarion

is

described

as

beingdue o habituat ionand sconsideredo

be

outside

of

the

subject 's o luntary

control .

In

out l ine.

he

habituat ion

hypothesis

grees

wi th

the not ions

descr ibed

n

the ntroduct ion

to

explain

the

spacing

effect.

There

it

w,ils

su_ugested

hat

a massed

epet i t ion

esul ts

n

the

subject

remember ing

he

solut ion

to a

problem

rather

han

const ruct ins

hat

solu-

t ion. This

remember ing

f

the

solut ion pro-

duces

poorer

retent ion

so

the locus

of

the

spacing

effect

is in

the

re_gistration

f rhe

second

resentat ion.

he

conclusion

hat

the

re-gistration

f the later

presentation

s

defi-

cient

s

compel led

y the

inding n

the

presenr

exper iments

f an

ubsolure

ebi l i tat ing

effect

of

repet i t ion

when

readine

the

solut ion

immediate ly

receded

resentat ion

ia

prob-

lem.

Further.

he

nf luence

i having

ead

he

solut ion

is

not

seen

as

being

opt ional:

t is

near ly

impossib le

to

be

uninl luenced

by

having

ust

read

the

solut ion

when

one is

so lv ing prob lem.

Althou-eh

agreewi th

cla ims

of

the habi-

tuat ion

hypothesis.

har

s habituat ion' l

hat

is,

what processes

re nvolved

n

habituat ion' l

I t

may

be

possib le

o

descr ibe

abituarion

by

appealin_u

o notions

hat

have

been

used

here

to

explain

the

effect

of

spacing

eper i t ions.

Perhaps

a

habituated

st imulus

is

one

for

which

an encoding

an be

remembered

ather

than

const ructed.

h is

vievr

oi habituat ion

contrasts with

a view'

recently proposed

b_r-

Wagner

l976 l .

Wagner

la ims

har

n 'hen

an

event

is

al ready

represented

n

short - rern- l

memory. ur theroccurrencesf that eventare

rendered

ess

f fect ive

han

they would

other-

wise

be.

A

simi lar

h,vpothesis

bour

the

importance

f

short - term

memory was

ested

in

Exper iment

of the

present

nvest igat ion.

There

t

was

concluded

hat

the

solut ion

o a

problem

did

not have

o

reside

n

short-term

memorv

o inf luence

he

solving

of

the

prob-

lem:

al l that

appeared

ecessary

as

har

t l te

solut ion

o

rhe

problem

ould

be

effort lessl1,

re t r ieved.

imi ia r l y

or

hab i tuar ion .

t

ma\

only

be

important

that

a

pr ior

encoding

oi

an

event

s ret r ieved

o

an

encodin_e

eed

not

be

const ructed.

h is

assumes

hat i t

is

the

necessi ty

f

const ruct ion

hat

_sives

ise

o the

or ient in_e

esponse

observed

n

studies

of

hab i tua t ion .

A

ser ies

of

exper iments

bv

Waugh

and

Norman

( l96 t t )

may

be re le ran t

ro

under -


13/19

standing

he effects

f spacing epetitions

and

the

processes

nderlying

habituation. Waugh

and Norman were nterested n specifying he

nature

of

an event

that

would

displace an

earlier

event from

short-term

memory.

The

results

of their

experiments

revealed that

a

new

and

unpredictable

event

would

displace

an

earlier event;

however,

a

repetition

of a

recently

presented

event

would

not.

If

we

identify

short-term

memory with

conscious-

ness,

t appears

hat

the

processing

f a

massed

repetit ion

is

automatic

in

that

it

does not

heavily

involve

consciousness.

Combining

this

piece

with

arguments

made

earlier

we

arrive

at the following picture:

Presentation

of

an event whosesolution or

encoding

can be

easily

remembered

does not

give

rise

to an

orienting

response

or heavily

involve

con-

sciousness;

resentation

f su ch an

event

will

also have

itt le impact

on

later

retention.

Th e

necessity

f construction,

n

contrast,

givds

rise to

an

orienting

response,

nvolves

con-

sciousness

o

a,qreater

degree,

nd

produces

a

substantial

effect

on later

retention

perform-

ance.The

spacing

of

repetitions

has its

effect

by determinin_e

hether

a solution

or encod-

ing

can be

remembered

r must

becontructed.

The

Generulitv

of'

Elfects

of

Constructiotl

Effects

an

be

ound

using

manipulations

n

addition

to

those

employed

in

the

present

experiment

and,

therefore.

the

speculation

about

different

modes

of

solvinga

problem

or

responding

becomes

more

interesting.

Before

-eoing

on to

deal

with

some

negative

effects

f

remembering

a

solution

or encoding,

on e

positive

effect

will

be cited.

A consistenr

ind-

ing reported

n

many

reaction-time

tudies s

that

the response

o

an event

hat is

repeated

s

quicker

than

the response

o

an

event that

occurred earlier but was not the last one to

occur.

Bertelson

(1963)

has proposed

rhat

when

a stimulus

is

presented

a subject

first

checks

memory

to

see

f the

presented

timulus

is

the

same

as the

one that

immediately

preceded

t. If

the

stimulus

is

the

same, he

subject

makes

the

same

response

as he

di d

RETENTION

EFFECTS

OF

SOLVING VS

REMEMBERING

661

previously:

f it is

not the

same.

he subjecr as

to retrieve

a response

hat is

appropriate

to the

presented timulus.The retrievalof a response

takes

additional

t ime

so responding s

more

rapid

when

the retrieval s

not necessary;

ha t

is,

when

the

subject

can simply

_eive

he

same

response

as

was

_eiven

o

the immediately

preceding

stimulus. Bertelson's

distinction

between repeating

a responsevs

retrieving

a

response

s

essentially

he

same as the

dis-

tinction

that has

been drawn

here betrveen

remembering

a

solution

vs

constructin*s

solution.

Repeating

a solution

is

more efficient

than s

constructing

one n

that repetit ion

of a

solution

can

be done faster

and.

perhaps.

with

less involvement

of consciousness.

Further.

there s

some

evidence

Keele,

1969) hat

ca n

be

interpreted

as

showing

hat the

repetit ion

effect ound in

reaction-time

studies. ike

the

effects ound

in

Experiment

2, are not

limited

to

short-term

memory.

Slamecka

Note

1)has eported

esults

ha t

are

similar

to those ound

here with

the

once-

presented

tems.

Slamecka

ound

that

generat-

ing

a response

o an item

(e.g.,

rhyme

or an

associate

of the

presented

item)

produced

better

later retention

than

did reading

th e

same response.One factor that differentiates

reading

a response

from

contructing

a re-

sponse is

that

the task

of constructing

a

response

s a

more

diff icult

one.

Severalexperi-

mentshave

shown

hat

a

diff icult nital

task s

associated

with

high

levels

oi retention.

l l lustrations

of the

relation

berween

he

diff i-

culty

o[ an init ial

retrieual

and

subsequent

retention

evel

have

been

provided

by

Gotz

and Jacoby

(197a)

and

Whirten

and Bjork

(1977)

among

others.

A

parallel

series

of

demonstrations

as

elated

he

diff iculty

of an

init ial

decision

o

subsequent

erention

evel.

For example, in one experiment by Jacoby.

Craik, and

Be_eg

in

preparation)subjects

ere

required

to

specify

which

word

in

a

pair

referred

to the

lar_qer

bject:

later

retention

was

higherwhen

members

f a

pair

were

close

in

size

flea-ant)

ather han

highly

discrepant

in

size

(flea-elephant).

Aubel

and

Franks


14/19

662

LARRY

L.

(1978)have

emonstrated

hat the

diff icultyof

comprehension

influences ater retention.

It

was found that requiring additional

effort

toward comprehension

of a

sentence en-

hanced

recall

so

long as

the sentence

was

eventually understood.

There are some

reasons to suggest

that

repeatingan

item a large

number of

times

has

1978 on interpreting the effects of repetition - solving a problem versus remembering a solution

Documents