DOCUMENT RESUME

ED 037 349 SE 008 085

AUTHOR Klein, Carol A.TITLE Differences in Science Concepts Held by Children

From Three Social-Economic Levels.INSTITUTION 2004 Randolph St.; Saint Catherine Coll.; St. Paul,

Minn. 55116PUB DATE Mar 70NOTE 16p.; Paper presented at Annual Meeting of the

National Association for Research in ScienceTeaching (43rd, Minneapolis, Minne., March 5-8, 1970)

EDRS PRICE

ABSTRACT

EDFS Price MF-$0.25 HC-$0.90*Concept Formation, *Elementary School Science,*Scientific Concepts, *Socioeconomic Status,*Student Characteristics

Fifteen elementary schools (310 fourth-grade pupils)used in this study were classified as serving high, middle, or lowsocial-economic groups on the basis of information from the cityoffices and the administration' of the metropolitan school districtparticipating in the study. The children in all of the schools usedthe same text and had the same amount of instruction per week.Teachers with little experience and those with many years ofexperience were distributed among the schools. Based on conceptsselected from the district's basic text, fifteen questions wereselected for the test. In addition to answering the questions, thestudents were asked to suggest a way to find out the answer to eachquestion whether or not they knew the answer. (Significantdifferences at the .01 level were found between the means of allthree groups on the I.Q. test and raw test scores on the test ofscientific understanding.) Significant differences were also foundbetween social-economic groups when raw test score means wereadjusted for I.Q. differences. (BR)

U.S. DEPIRMII1 Of MIDI EDUCTION t WEILIN

OFFICE Of EDUCATION

TINS DKUNENT NIS SEEM IEPIODUCED EXACTLY AS WEIRD FROM THE

MON OC 014ANIZATION mum IL POINTS Of VIEW OR OPINIONS

Differences In Science Concepts STATED DO NOT mummy REPIERNI OFFICIAL ME OF EDUCATION

POSITION OR POLICY.

47% Held 2y Children From Three Social-Economic Levels

4N. Carol A. Klein

reN College of St. Catherine, St. Paul, Minnesota

CD

LLI

Science education has now become more than an incidental

part of the elementary school curriculum. Many new textbooks

and curricular materials have been developed or are in planning

stages. Only a few of these new programs, however, are concerned

specifically with materials for the educationally disadvantaged.

Children are disadvantaged because of low family income, by

being a member of a minority group or race, by living in a

ghetto or inner city, or by moving with migrant working parents.

Any of these disadvantaged or combination of them may result in

educational problems.

Although programs such as "Head Start" have been implemented

because there is evidence that social-economic conditions influence

a child's readiness for learning and success in school, once the

child is in the formal school situation, any differences resulting

from social-economic influences are largely ignored, and the same

curriculum often is used for all of the children in the city.

Related Studies

The literature reviewed for this study was limlced to re-

search reported in the last twenty-five years, because the

nature of science teaching and the sources of information

children have available are much different than the nature

study oriented science before 1945. Much of the literature

examined included little or no statistical treatment of the data.

This wr=s due in part to the nature the studies, and in some

cses, to the simple failure to report important data. (44.CitrUaleS

that were characterized-by clear d .mn ana 6. at were concerned

with assessing the science concepts of children, or the sources

of information children use in science, were included in the

literature study.

In studies where social-economic grouping was considered,

AlmyI

found that children in the lower social- economic classes

follow the developmental steps postulated by pia ;et more slowly

than children from upper or middle social-economic groups. Dart

and Pradha2n found considerable differences in children's

concepts of science and the nature of knowledgewhen they compared

the children in Hawaii with three sub-cultural groups in Uepal.

This study was still in the pilot stage and no statistical data

were reported.

4Investigations by Anderson,

3McCollum, Brown,

5.,:mpt,

6and

Inbody7 provided information about science concept development in

children.

The sources of information children use in science were in7

11vestigated by Schenke,

oYoung,9 Bergen,

10and Kuse. The

methods used to determine the sources of information that children

used varied, as did the results. In some cases almost half of the

science information possessed by children was attributed to ob-

servation and experimentation. In other studies, books and parents

or teachers were most frecuent sources of information.

. *)

. ,. , ((I .f :0. $1 4.)I 1 :3 .. i 31 rr.1;; 0 .)

A I 4.) gi r: iO c 17.3 047 11 s.: ; 1 g;:1 ; 't ri,!-Ia) o D o u)ri H (..) ,c-1 ti3

':4 ::. I 0 4-1i

::' 0 () 03 n3i d .1)0 0 In i-i

rrji-) 14 11 0ii *.? 0

1.1 0 ..% 4-) u)rd 0 0 riO S:1 (J) 01 J1r1 0 e.) c-i -)0 0g: rd C) gl

4-1 0 ;:i .4.- I riC) .1.) r : riid s::: (r) 0 .C:

4.14.?0 VI C) H

4- Irl P4 0 /.1 ,.10 0 1:1 C.) 0) 00 01 0 3-1. r I (!):4. i..4 cn P40 u) i --I (1)Pi d Si 0

4:14.41

:.6) 0 6) iC

41) 4 0 1141 rd V) 0rc..4 4-1 (r)

4 I 4-i4 )

c1,i 0 4-)

1 4) .C.:

0.3 () ( al

X t) 4), gl 4)r.1 t : 0,, Cl)(') rl t.' ..4t . rd

.4.3i H

E i f,

S C) C)

r.-I U)r. -1 )

I f: I

$ I ,4 51 0 C)

-)4

... 4-)C) 4.4 04- I 0

.4 ,..'.. 0gl

1.

Lf)Irl 0)

0 ci 0 0)@

Cl r;S .1 u) H U)0 ,e) C).1 AI *pi

Id In rdC) N (A 7iO 0 C) 4 ):i u) Pi V)

0+3

F.1l) 4.1 4.31 0 (1.1

1.1 Cl)tl LO

C)) r)1i I3ha

$ rd (c)4ci $1r-i d rdi -P H Er:e41 U) .3 i)C) $+ p f 140 0 0

4-4 1,3 0.4.-1 g-I

C-

:*4 4.41 0-) o

e.) ...0 P.O 3-4 4)'4i ri t))0 in Al

3.4

0 4-) 0,0 rd 1.).

4-) ,c+ ..00 31 4) 4-)d ri 0 .41E i0 rci

,:4.) q) Cr) rd

14 A-1 Cl)LI .4-) .I.3IV. q-4 d.:- ri-4 s1 4-1ri 4-1 C),`=-, ici 0rcj rd rn

t cl fn4

1-)

r)P4 01 CJ

U) II0 u) co

u) $1 +3 )1, ri b0 Pi 0..0 0 .4-I4-1 C) 0 4-)

. r I t:31 V) (I)4-4 0) Ho o :1 0

P.' CI 04)) 0 0 .4.I)) 0 (,) 0 C.)C) 0 4.:' +3 (1)P4 I ni$., 1 .r1 u) 4-1ri 0 $-4 0Pt .ri tn C)

()0 0 r

f::rd in 0

.( 1 N 0 0 4-4E-4 .4-) el 4-)0 0 Ci

6) 6) rd Cl)$.I 4-.1 41 r1,r; 0) i-i 04.) u) 4 -i CI)

'Ticccn 0) 4.,.)0 :-.-, H ..

RI 1(.ti

.. '..' :

+0 0 ,'L.: 'C'l () Cn-1.3 1.4 4 ..C.: 4. 4.

c..Q14

4 a.00 'ez: 4-4uti 'Titd

..1)0 C) v) ,s1 (I) 41,) . 40) 0 ,-- 0 u) 4-) 4 ) I:4 XI i 4 4.)H 0 44 SI r-i 4) 4) in 00

H . d4)) n)

$. i . 4,:.1

0 )- 0) 0 Cl) td .a .J4..) 4 3'1 4-) 4f2 Al $.-i Aj rd ri 'Li 13---'0 4, .4-4

(13 ri 4)4-1 p 4.1 c;.1 114 t5 Cl) H

4-04 ,4'11$4:, Adi rE11)

C)U] al 31 P4 0 1r -{ 0 4-) 0 e3 .3-I (4) ti)

P40 Cl) ra Pi ,0 o 0 o 4 -t f.1) r..I0 H 0 Q 4) 0 .4 0 0 41 .4-1. I.:- H4.-4 3-4 ,. 1-4 0 n3 ni 4;3- d,r1 rd

C) d 0 4) Cl) 4-1 6) $1 <1.3 43.) 0Cl) 'H 0 0) H 0 :- Pt 0 QJ 4-) -) cf) C) rd4' 1..1 Ei 0 SI 4-1 ri $.4 X. 14 0 (Ii-4.4 n 4.)

al0

0) rc: rci1-ci

tnis4-)ni PI

:73 rid 0 0 0 ,31

ni ,.c.:._Ei .-`.-2

0 0) 0 0) 0 19 . 0 -00 $.1

P4+3 110 +3 CO cr...1 rir:i.1 riii e.°C.1

) <I 0 ..--1g.4. .ri (.) 0 Cr)(1) 0 0 PI o@ ,-,-, 1-1 rd cd 0 4.) 4.-1 (d El X 141

,C1 cc-4 4-40 0 1-1 0 0) :',. c3 co 0 ID0 V. gl g-i 4.

F.; PtH 0) 0 ri (1) . 4 -t 00 s.tt:20

r 4

14 4 -IPI 0 .4-I0 i 0 .r: t. a 0 t:I: 4 ) : IYI I-F..? i--°(1).1 4..4 31 4-) 73 0d N

0 .4,i'. al 0 ro ((3 A (,) U)

.13 0 Cl) cd 1-,e ) Cr ..1 P 4 1> u i rd 14- $-i ,.00 u) .i 1 $4 rd 4-) 0 0 0 Pi CJ Cf .....N u) t.) ::$ C.) $.1 fl **-1 G.)I 0

4 o -P ,ci.1) 0) (d P 1 0 .1 3 H 0 0.1 i>b1.1 (11 40 ei to c1 c3

01 4) rd0 cn ri Pi0 gl 4 44 d cri ).

'CA 0) c3 03-.1 0 :-.34) P. i c-1....1

ttei ,ec):: .ri0 .( : PI 'CS rd 0 cd 0 0 el 31

I ..4.-1 4-3 ..c.3 4-) .-.3 ::-1 t)f) 4-) oiIN t f_.4

03 0 6.) r I es) o .4 ) ... 4) o)H H gl 51 0 1-4 Cft 0 0 ..C1u) 0 .40 E...31

0 0 4..1 -4 3 0(.3 u)

e) *r Ir.I I.> " >) .:ii -1 .5:(14 (13 u) 4 ) r-0 0 (3 C)) ,tp 0) 4) .... -0H P c:: ri . (I) 4.) ..C.: (')

.-

(Cf.)

0 :3 rci 1-1 ICJ 'Ti PI ..5::: ,d .0 C1 .,11.1 C)4-) 0 4:1 .4) 0 0 0 (1) 0 0 ri -t(I) 0) $4 lA.1 el 0 rd 0 0 0 0Si Li) P4 0 ', Cl)

4 .r4 (4.4N HPi )4

0 C) . kl (.) .51gi Cl) cf) ,1 0 H 10 ;:i 0 4-.) H 0(1) .:-. 0 0) 0 6) C) 0 Pt r-i 1 0 d0.) -,-4 o *4-4 1 t.: 0 H 0 r(1 0 0g: ..t1 .4-)

is ',..,-, 0 .1.1 4.) r-i :10 XI 4-4e) 0 (.1

°C) rti4., ,ci 0 4-4 1-1 03 $-4 id al - Cri

(13.ri I.'S (1 ri?A .-1.9 .1-9

tn Ii) 41 (r)4.4 4 ) 0 e01-4 P4 0 r- 1:0 ri 0 0u) c) o e) cd 0 0 4-4 c) 0 C)0 0) > -P u) oj; : 131.) : CI ..; --)4 41 --)4

0 0 t.:CI 6) I 4 ) C) (4 0)i ,s. 1

r-(ll.,..i 4,-/ 0 01 d il3 4 . ; . ) i°1-1 : 49

31 34 t 14.3 0 14 H 3").4 4 Pt 34 r(14-4 4) 1:0 0 0 (1) 0 4-1 0 0 C.: 0$-i () 0 u) Si 0 P -P C) id ri 0 P4 0 Pi0 0 .s.1 H H $ ,c; r.! ;--i t.! 4.11

I Ael,i In -i.) rei u) t..0 .i-i ILI 0 el 0 t4i 6 d c.)

-cience-cuesti-onssea on concepts selected from the system's

basic text were tested in a suburban school district (co=prisecl of

children from middle and lower social-economic groups) to d,_te=inu

"1"

--A 1,-uneir clarity cliscrimina4ing power. In addition to ansverin-,

the fifteen cuestions finally selected for the test, the stude-_ts

were askea to suggest a way to find out the answer to es.ch question

whether or not they knew the answer. The following is an example of

the questions in the test:

12. 'nhich of these three animals is the grown up? 1, 2 or 3?,

Why did you choose this one?

if you did not know which was the grown-up, how could you

find out?

(This cuestion is accompanied by a picture of a caterpillar,

a cocoon and a butterfly)

In the May, 1908, pilot study, the test was administered in-

dividually by the researcher to sixty third grade children. Their

response to the questions were tape recorded. Later, forty of

these children took the test in written form. The questions were

read aloud to minimize the possible consecuenaes of different

reading abilities. The means and standard deviations for the

written and taped test scores were very similar and correlated

.87 to .90. The test-retest reliability was, therefore, at these

values. Acomparison of interview and written tests can be made from

the data in Table T.

-4.1O_Lt. r.

Means and Stanaard Deviations

gor Written an3 Interview Scor,..:s in Pilot Stud

Social-Economic Group Means

Interview Written interview

LowSchool A 4.10 3.90 1.72 n

LowSchool 3 3.40 3.30 1.17 1.15

MiddleSchool A 7.80 7.70 1.87 2.11

HighSchool A 8.4o 8.5o 1.66 217

Test Validity and Reliability

Test validity has many meanings. Ebel12

suggests that an important

consideration in test validity is the purpose for which the test is

used and the group with which it is used. The test in this study was

designed to determine if there were differences between children's

understanding of selected science concepts and the methods of verifi-

cation when social-economic groups were considered. The questions

used for the purpose of determining the understanding of the

crildren were based on the concepts included in the science text-

book which was used by all groups The children were asked what they

could do to find out answers to the questions asked. This basis. for

cues titan design and selection was believed to satisfy Ebel's

definition of validity.

An individual test item should be answered.correctly more often

by students achieving a higher overall score than by students with

a low overall score. It should discriminate between thr-

compared. Achi-saua-e test independence on th.

in the Pilot Study showed significant dil'gerences, at the .C; leval,

between the un_et. grow S for 11 o the 15 titestesems. P'he hifnJ. .

social-economic group answered all items correctly more often than

the low group.

Atest of internal consistency was also used on the

view data. Using Hoyt's formula13

an internal consistency of

was obtained.

The results obtained in the pilot study were believed to

justify using the written form of the test for the main study.

In the fall of 1968, the written test was administered by the

teachers in the same manner as the pilot study writ ten form,

providing data from 310 fourth grade students in fifteen classes

in nine schools

Ma n Study P..u1.

The five null hypotheses considered in the main study were

concerned with differences between the three social-economic groups

in I.Q. mean scores, raw scores on the test of scientific under-

standing, the means of the raw scores on the test of scientific

understanding when the means were adjusted for I.Q. differences,

the methods used to find out answers to auestions and an item

analysis for the science test.

no rirc....1 On The I.Q. scores were obtained for each

child in the main study from the school district participating in

the study. The Lc-n.e-Tro-e-ndie. ,n..l.i-erce test had been given

to the students one month before the science test. A one-way

,-was used to t1le null nvpothes:_s of no

sisni,"icant difference between the mean I.Q. scol-cs

tr,rouns. 1.1,e null :""00÷S-S 1-.'ec'ed at the .01 levc:.

Student-Newman-Keuls test established that ea-el, group re= was

significantly different from each other group mean. A sum:Lary

is shown in Table IT.

Analysis of I.Q. Test Scores by Social-Econcni c

Social-Economic Group I.Q.

High Middle Low

No. Students 78 97 135

Mean 116.13 107.41 99.72

Stand. Dcv. 11.39 13.04 11.80

99(2,200) = 4.71

One-Way Analysis of Variance

F value = 45.97 Reject Hypothesis

Student id-,Iewman-Kuels

Significant Diffe-^enne. Between Means by Levels

Levels

Middl

Low

Middle Low

...%

The second hypothesis to be tested was

that there was no difference in the mean level of scientific

understandin;., of children from the three groups. A one-way

analysis of variance test on the raw scores or the ties grou-es

and the Student-Newman-Keuls test to determine the significance

of the difference between means was conducted. The null hypothesis

was rejected at the .01 level and each mean was found to be signi-

ficantly different from each other group mean. Table III summari2.es

these results.

Table TTT

Analysis of Mean Score on Science Test by Levels

Social- Group Means

High Middle Low

No. Students 0,,

1 97 =35

can 7.04 0.14,- 1

2.71

Stand. Dev. 1.94 1.75 1.72

(2,200) =4.71

99

One-Way Analysis of Variance,

F value=179.83 Reject Hypothesis

Level

Hi:h

Low

Student-Newm,,n-Keuls

Sign -,-ificant Differenc Between Means by Levels

High Middle Low

ti

9

HvPzIthosiz Myee. of covariance 1.13cd to tc.12:..

the third null hypothesis. When ra w score means on the test fu:

scientific understanding were adjusted for I.Q. diffc-r.e:_ces,

null hypothesis of no difference in the mean level of s:ientifi2

understanding of the children for the three groups was reject,,d

at the .01 level. The Student-Newman-Keuls test showed that the

low social-economic group mean was significantly lower than either

t. middle or high group mean but that there was no significante% "o- 6

difference between the middle and high group means. A s.LI zimary is

shown in Table IV.

Table IV

Raw Test Scores Adjusted for I.Q. Differences

Social-EconomicGroup

High

Adjusted Mean 6.29

Grand Mean = 4.87 Test Score

6.06

Low

3.21

F99 (2,200)=4.71

Covariance Analysis

F Value = 114.61 Reject Hy pot

Student-Newman-Keuls

Significant Difference Between Means by Levels

Level High Middle

T-ria0cr. l

Middle

LOV

Lvrotl:ot-.is h. to be tested was

t:lar .i! is no aifference in tsar the students of the three

social-economic .1-roups would use to verify obtain

to cuestions about their scientific unde-rstandingc,. Since all

answers were considered co-rrect, no test of stntistical sirnifi-

cance could be used. The per cent of answers given in cz,z.h category

is found in Table V. Although the children in the low and high

cq-ouns used books as a source of information about -...-:

of the time, their actual answers were much different.

De:: cent

In ...re

lay group, "dictiona,7" was usually specified while the high

social-economic group often indicated a "science book about

weather- 2 or "a book about sTa:ce" or some more specif4c source.

TablA, V

Methods CYldren Statd to Ve-rift' or Find Out Answers

to Questions of Sciertil'ic Understanding

Social - Economic Group

t hod Middle

Ask Parent 2.13 1.38 2.66

Ask Teacher .52 .06 .29

Avl- r-1,^-_0_ ,..J..,. 3 3i 3.16 4 0,--...,2

3ook-:ewspaper 14.28 5.22 13.98

Radio-TV .94 .97 .69

Obse,-ve-Expel-iment 74.87 78.90 56.74

1To Answer 3.42 10.31 21.29

:,,,abers indicate of times stated

Hyl-othesis Five. The fifth hypothesis of no di22, rence

in the number of correct answers riven by the three groups to

the individual items on the test of scientific understanding was

rejected for 13 of the 15 itens at the .05 level. The chi- souare

test of independence was used. The two questions failiz below

the chi-square value were, however, answered correctly more often

by the children in the high and middle groups.

rt...=1-

1. Is there a significant

difference between the

Summary of

Questions and Results

Tested by

One-way analysis

of variance

can I.Q.'s of the three Newman-Keuls Test

social-economic groups?

2. Is there a significant One-way analysis

difference between the of variance

mean scores on the test Newman-Keuls Test

of scientific understand-

ing when only raw scores

are considered?

3. Is there a significant Co-variance analysis

difference between the Newman-Keuls Test

mean scores on the test

of scientific understand-

ing when the raw scores

are adjusted for I.Q.

differences?

Results

S,c0. ri. cant, difference

at .01 level between each

group.X=High 116.13

'Middle 107.41 Tow 99.72

See Table II

Significant difference at

.01 level between each

group.X=High 7.01

Middle 6.14 Low 2.71

See Table III

Significant difference

at .01 level between low

and high, 3.03-6.29;

low and middle, 3.03-

6.05.

See Table IV

0cstion

4. Are there differences

in the kinds of answers

suggested by the three

3e cd ,r

Differences by percent

groups to the questions 1.--ast by lov

"How could you find out?"

5. Are there differences in Chi-square analysis

;,ne number of correct

responses given by the

three groups to individual

test items?

group. Low often

no answers suggest ::d.

See Table V

Significant at

.05 level for 13

of 15 iten',.

had least correct

on each of the

J.J.ateen.

Conclusions

If differences in levels of understanding are influenced by social-

economic factors, it would seem necessary, to this writer, to consider

this when planning a curriculum for elementary science education.

Metropolitan school districts using the same curriculum materials in

schools in a city that has a neighborhood school system, as well as many

groups planning new elementary science curriculum materials, are in most

cases ignoring the influence of social-economic factors on children's

understanding.

If science education is to be concerned with the process of sci-!nce,

experiment and observation, as well as the concepts of science, then

sonic changes may be necessary in the curriculum. In the science test

used in this study, the answers to all of the ouestions could have

been determined by observation or experimentation. This methof, of

finding out answers was suggested 56 percent of the time by the

children in the low social-economic group compared to about 75

percent for the middle and high grouns. Acurriculum planned for

children in the low social-economic group may need to include many

opportunities for solving problems by observation and experimentation

if development of this ability is accepted as an objective of science

education.

In the pilot study of this research, the children were given a

test of scientific understanding first by an interview method and

two weeks later by written means. The test questions were read to

the students taking the written form of the test so that wrong answers

or failure to answer would be due to inadequacy of science background

or knowledge rather than reading problems. There was very close agree-

ment between the means and standard deviations for the interview and

written tests and a high correlation of .87 to .90 between raw scores

on the written and interview tests. It is possible that this method

could be used in other research studies where the investigator wanted

to test a large group of students but was concerned with the disadvar.tage

some children would have on a written test because of their poor reading

ine be substituted satisfactorily fa-theability. m. *4-4. 4- ,4 4-*4- 4- 4-*

interview method if the written test was read to the students.

Tais method of interview -uent and written retest may also

provide one means o-fs establ4s.linz when the standard

test-retest method is not desirable.

Children in all of the groups, had the least correct answers

to euestions about the motion of the sun or on the influence of heat

on the contraction and expansion of mate r ials. The Questions concern7-d

with living things were answered correctly more often bv -.;he children in

all of the groups. ine metamorphosis of a butterfly was much more

frequently understood by children than the motion of the earth around

the sun, but if these concepts are included in the curriculum, as

they were in this study, then better ways must be devised to help

the children develop concepts of more abstract phenomena.

S,rno -sis

A test constructed to determine selected science concepts of

children and the means they suggest to find out the answers to these

Questions was given to children in three social-economic levels. In

- . 4. J..',Lie D110%, study the results the interview and re-ues. wr itten methods

were thought to justify using une written Jorm Oi the test in the main

study involving 310 fourth grade children.

Significant differences at the .01 level were found between the

means of all three groups in the I.Q. test an ray test scores on the

test of scientific understanding; between the low and middle and low

and high social-economic groups when raw test score means were ad-

justed for I.Q. differences.

- 15 -

If differences in levels of understanding science are as-

sociated with social-economic factors, using the same text and

may not be an effective teachinc,0 method. 1;ore CD-

portunities for experiment and observation are needed.

1. Alny, nillie, "hinkinf:. :P.a. York: Columbia

University Teachers Colle-:c ?ress, 196 7.

r,art. Francis E. an;; P-,nn-, L. "Cro,.s Cultur-.1

Teach'n- of Science". " c c7 _reoruary 1)),

AnZerson, 2anald D., "Childrens Ability to Pormulate MentalModels to Explain Natural Phenomena," Unpublished DoctoralThesis, University of Wisconsin, 1964.

4. McCollum, Clifford B., "A Technicue For Studying The Maturityof Elementary School Children in Science," Science Education,April, 1952, 36 168 -175.

5. Brown, Stanley B., "Science information and Attitudes c., sessed

by Selected Elementary School Pupils." Science Education,February 1955, 39, 57-59.

O. Haupt, George W., "Concepts Gf Magnetism Held by ElementarySchool Children." Science Education, April 1952, 36, 162-163.

7. Inbody, Donald, "Children's Understanding of Natural Phenomena."Science Education, April 1963, 47, 270-278.

8 Schenl-e, Lahron H., "In4'ormtion Sources Children Use."Science Education, April 1956, 40, 232-237.

9. Young, Doris A., "Atomic Energy Concepts of Children in Thirdand Sixth Grades." School Science and Mathematics, October 1958,

58, 535-539.

10. Bergen, Catharine', Some Sources of Children's Science Informatior.New York: Columbia University Teachers College Press, 1943.

11. .'fuse, Hildegard. R., "A Survey of the Sources and Extent of PrimaryGrade Children's Concepts of Elementary Astronomy." Unpublished

Doctoral Thesis, University of Colorado, 1963.

12. Ebel, Robert, :-feasurirK; Educational Achievement, Englewood Cliffs,

New Jersey: Prentice Hall, 196-5.

13. Hoyt, Cyril J., "Note on a Simplified Method of Computing Test

Pv.liability." Educational and Psychological Measurement,January 1941, 1, 93-95.