You are on page 1of 32

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233463029

Structural patterns in students' conceptions in mechanics

Article  in  International Journal of Science Education · July 2010


DOI: 10.1080/095006999290228

CITATIONS READS

21 47

1 author:

José María Oliva


Universidad de Cádiz
128 PUBLICATIONS   901 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

El enfoque reflexivo en la educación científico-técnica y la formación docente View project

Desarrollo y evaluación de competencias científicas mediante enfoques de enseñanza en contexto y de modelización. Estudios de caso. (EDU2013-41952-P) View
project

All content following this page was uploaded by José María Oliva on 11 June 2014.

The user has requested enhancement of the downloaded file.


INT. J. SCI. EDUC., 1999, VOL. 21, NO. 9, 903– 920

RESEARCH REPORT

Structural patterns in students’ conceptions in


mechanics

José Ma Oliva, Centro de Profesorado de Cádiz, Spain

This study reports the existence of patterns of reasoning that are common to conceptions in different
areas of mechanics. The level of generality of these patterns is also studied, showing their relationship to
the general frameworks of formal thought. The chosen sample was made up of 200 grade 9 and 10 pupils
from three state schools. Questionnaires about conceptions and the TOLT (Test of Logical Thinking)
were used, as well as individual interviews of a small sample of pupils. The results revealed the existence
of certain causal frameworks common to different ideas with an intermediate generality level which lies
between that proposed by Piaget’ s stage theory and the more fragmentary approach of alternative
conceptions. Several conclusions about conceptual change models are derived from these results.

Introduction
Studies on alternative conceptions in science education have adopted one of two
opposing standpoints. On one hand, there is the descriptive approach which con-
siders the student’s conceptions to be a set of specific notions formed indepen-
dently of each other and closely linked to the content and context of each task in
hand. The structuralist approach, on the other hand, maintains that the different
notions stem from more general frameworks of reasoning. Piaget is the most radi-
cal supporter of the idea of structuralization. He holds that the concepts and
notions formed by adolescents in specific domains depend on general structures
of reasoning that are independent of the content.
The present tendency in research is of evolution towards a halfway point
between the two extremes. This has been brought about by two types of conclu-
sions:

• Although the conceptions used are usually dependent on the context of the
tasks, a statistically significant predisposition in each individual to give
similar reasons on different occasions has been detected (Engel-Clough
and Driver 1986, Jiménez-Aleixandre and Fernández 1989, Licht and
Thijs 1990, Oliva 1994);
• Even when the level of generality for knowledge postulated in Piaget’ s
theory of the stages does not seem consistent with the experimental data
obtained, a certain regularity in causal thought through the wide range of
catalogued conceptions in diverse fields has been observed (Viennot 1985,
Anderssson 1986, Monk 1995).

International Journal of Science Education ISSN 0950–0693 print/ISSN 1464–5289 online Ñ 1999 Taylor & Francis Ltd
http://www.tandf.co.uk/JNLS/sed.htm
http://www.taylorandfrancis.com/JNLS/sed.htm
Structural patterns in students’ conceptions in mechanics

José .Mª. Oliva. Centro de Profesorado de Cádiz. Spain

This study reports the existence of patterns of reasoning that are common to conceptions in different areas of
mechanics. The level of generality of these patterns is also studied, showing their relationship to the general
frameworks of formal thought. The chosen sample was made up of 200 grade 9 and 10 pupils from three
state schools. Questionnaires about conceptions and theTOLT (Test of Logical Thinking) were used as well
as individual interviews of a small sample of pupils. The results revealed the existence of certain causal
frameworks common to different ideas with an intermediate generality level which lies between that
proposed by Piaget´s stage theory and the more fragmentary approach of alternative conceptions. Several
conclusions about conceptual change models are derived from these results.

Introduction

Studies on alternative conceptions in science education have adopted one of two opposing
standpoints. On one hand, there is the descriptive approach which considers the student´s
conceptions to be a set of specific notions formed independently of each other and closely
linked to the content and context of each task in hand. The structuralist approach, on the
other hand, maintains that the different notions stem from more general frameworks of
reasoning. Piaget is the most radical supporter of the idea of structuralization. He holds
that the concepts and notions formed by adolescents in specific domains depend on
general structures of reasoning that are independent of the content.
The present tendency in research is of evolution towards a halfway point between
the two extremes. This has been brought about by two types of conclusion:

 Although the conceptions used are usually dependent on the context of the tasks,
a statistically significant predisposition in each individual to give similar reasons
on different occasions have been detected (Engel Clough and Driver,1986;
Jiménez-Aleixandre and Fernández, 1989; Licht and Thijs, 1990; Oliva,1994).
 Even when the level of generality for knowledge postulated in Piaget’s theory of
the stages does not seem consistent with the experimental data obtained, a
certain regularity in causal thought through the wide range of catalogued
conceptions in diverse fields has been observed (Viennot, 1985; Andersson,
1986; Monk 1995).

The degree of structural organization in students´ conceptions in mechanics is analysed in


this study, from a viewpoint that is close to the second of the above conclusions.

Background

In pioneering study, Andersson (1986) proposed the same causal structure to explain
conceptions in diverse domains such as heat and temperature, electricity, optics or
mechanics. He analysed children’s responses to a large group of tasks and identified an
underlying common core. This common core consisted of a causal relationship by direct
contact between an agent and an object between which energy flows. The nearer the agent
is to the object, the greater the effect created.
On the other hand, there has been a tendency to analyse phenomena starting from
the changes that take place within them rather than from the states or situations of
equilibrium existing within them (Driver, Guesne and Tiberguien, 1985). For many
subjects, for example, the state of rest is not something that requires an explanation in
terms of forces. The process of acceleration or braking is not interpreted as being caused
by external factors but rather by certain tendencies that are naturally innate to bodies.
Acevedo (1990), Touger et al. (1995) y De Posada (1997) have found that some of
the explanations offered by pupils in science tasks are based on superficial personal
impressions, instead of being based on logical criteria that analyse the variables of the
problem. Moreover Watts and Taber (1996) have detected a predisposition to give reasons
such as ‘it’s natural’, ‘it’s normal’ or ‘it’s common-sense’ for physical phenomena. This
reasoning is based on natural tendencies intrinsic to bodies rather than logical inferences
based on an analysis of the elements that intervene in the phenomenon.
Another contribution of interest is to be found in a series of studies, that begin with
Closset (1983), which use a simple causal reasoning framework to explain the behaviour of
systems in different phenomena. In the case of Closset´s work, the framework is printed by
sequential reasoning and applies to the interpretation of electric circuits. The changes that
take place in this type of system only affect whatever is first met, the circuit being
understood as a succession of elements placed sequentially with no idea of the interactions
that take place within them. Similar explanations have also been found in tasks on
chemical equilibrium (Hackling and Garnett, 1985) and may be responsible for certain
difficulties in the comprehension of the notion of force as interaction (Maloney, 1984;
Licht and Thijs,1990).
In line with this way of thinking, it has been found that students tend to reduce the
number of variables that influence a phenomenon, in domains such as thermodynamics
(Rozier and Viennot,1991), the superposition of electric fields (Viennot and Rainson,
1992) or the propagation of mechanical signals (Maurines,1992). Together, the latter
two make up a more general framework of reasoning which may be called linear causal
reasoning. This framework reflects the tendency to use causal reasoning without
ramifications or retroaction and a difficulty in the interpretation of phenomena which
depend on several factors (Viennot and Kaminski, 1991).
On the other hand, Stavy and Tirosh (1996) revised a series of papers in which the
intuitive rule ‘the more of A... more of B’ is the common denominator. This is a direct
covariance rule which is normally applied when intuitive inferences about experiences are
carried out. The authors mention, amongst other things, the tendency to conceptualise
falling objects in terms of “the greater the weight, the faster the fall” which implies a
systematic and implicit association between the magnitude of a cause and its effects.
Finally, authors such as Pozo (Pozo et al, 1991) have classified the trends in a vast
catalogue of conceptions, identifying three mental forms or structures. These structures
may be expressed as continua or dimensions whose extremes represent implicit ways of
thinking in commonsense and scientific reasoning:

(i) Linear causality as opposed to interaction between systems.


(ii) Change and transformation as opposed to conservation and equilibrium.
(iii) Qualitative relationships as opposed to quantitative frameworks, such as the
frameworks of proportion, probability or correlation.

The above mentioned studies are of great theoretical and applied interest. Not only do they
offer a more systematic and comprehensible vision of the wide range of conceptions found,
but they also contribute to a new perspective on conceptual change. Most of the work
which has been done is of a theoretical nature or is based on a comparison of ideas
expressed at sample level. There is a lack of studies that evaluate the degree of structural
organization that is aimed at. One would like to see the same conclusions derived from
quantitative analyses comparing the replies given by each pupil in different cases. This is
the approach adopted in this study.

The study

The experimental part of this work attempts to establish whether or not reasoning
patterns exist which are common to the different conceptions described in the
bibliography. The field of knowledge chosen for the study is mechanics. Likewise, the
present study attempts to analyse the degree of generality of these patterns by studying
their relationship with the framework of formal thought. Part of the theoretical framework
that has served as a foundation for the development of this paper is based on Monk’s
proposal (1995). Monk identified five scientific principles which may be of use in the
study of pupils’ conceptions: structure, scale, orthogonality, statistical probability and
progression. This study takes four of these principles that are considered to be basic in the
planning of the methodology to be employed and in the interpretation of the data obtained.
The manner in which these principles have been used is similar to Monk’s proposal and
may be specified in the following way:

 Structure: The conceptions students have of scientific phenomena do not consist


of disperse entities but are structured into higher level frameworks.
 Scale: To delimit these common structures we must change the scale of analysis
and move on from studies of specific topics, in which only the peculiarities of
each conception are of interest, to studies of a wider range where interest lies in
the nexuses different conceptions have in common.
 Orthogonality: A large part of the diversity of conceptions and beliefs that
characterise a pupil´s conceptual ecology may be explained by means of a
combination of a small number of simpler reasoning models.
 Statistical probability: Depending on the viewpoint on which the work is based,
each conception is of a probabilistic nature for each subject, in the sense that:
(i) Conception assimilation in the cognitive structure is not a question of all or
nothing but of degree, and (ii) the use of each conception is not sure but
uncertain as it results from a complex process of decision-making among diverse
conceptions that co-exist and compete to be activated.

Viewed as a whole, the principles of structure and probability suggested by Monk lead to a
merely partial level of coherence for conceptions. The repetition of certain patterns or
causal frameworks is expected in different cases, yet not as systematically as described by
the scientific theories. This paper takes up this point of view which is close to that upheld
by Pozo in his work (Pozo et al, 1992), where ideas are considered to be of a widespread
nature, similar to that established in cognitive psychology for implicit theories of common
sense. Thus an intermediate stance between the position held in Piaget’s stage theory and
that proposed by the alternative conceptions approach is adopted.

Method

The study was carried out in two stages. During the first stage written
questionnaires were used for the analysis of the conceptions pupils held with regard to the
chosen topics. A paper and pencil test was employed to evaluate general logical-formal
thought skills. During the second stage a number of individual interviews were carried out
with a small sample of students.

Sample
Two hundred ninth and tenth grade students from three Spanish state schools participated
in the first stage. One school is in a provincial capital, another in a town near the latter and
the third is in a rural district. During the second stage, 11 ninth grade students from the
schools involved in the first stage took part.

The questionnaire on conceptions in mechanics

Two questionnaires, A and B, consisting of seven questions/problems were prepared as a


method of collecting information. The tasks in both questionnaires were quite similar
although the contexts of the situations were different. The student had to evaluate his or
her own viewpoint with regard to three items or statements that were interpretations or
predictions concerning the phenomenon being dealt with. A total of 42 items had to be
evaluated, 21 for each questionnaire. The choice of topics to be studied was made taking
into account the academic content of secondary school syllabuses, as well as those
alternative conceptions that are most studied in science teaching research (Hierrezuelo and
Montero, 1991). The topics dealt with in the questions and their corresponding
conceptions are shown in table 1.

[Insert table 1 about here]

The presentation format of the items was of the Likert type, with seven levels that
covered a series of replies that ranged from complete agreement (+3) to complete
disagreement (-3). The appendix at the end of this paper is an example of one of the
questions.
The proposed items were transcriptions of replies and reasoning contributed by
students in a previous study which was carried out employing open explanation
questionnaires (Oliva, 1994). The order of presentation of the questions was identical in
both questionnaires, although for each question the three corresponding items were ordered
in a different way. However, in order to facilitate the reading and interpretation of data in
this paper, the items have been placed in the same order in both cases. The tests were
carried out during the last weeks of class in the case of the grade 9 students and at the
beginning of the academic year in the case of the grade 10 students. Previous education in
mechanics had been limited to a brief introduction in seventh grade.
The questionnaires were handed out during normal class hours and, although no
time limits were imposed, all the questionnaires were completed in less than 30 minutes.
The students knew that they were participating in research to ascertain prior knowledge
and that the results would have no bearing on final school marks.

Test of general reasoning skills

All the pupils who took part in the first stage also completed a Spanish version of the Test
of logical thinking (TOLT) by Tobin and Capie (1981), the aim being to evaluate general
reasoning frameworks. This version of the test had been validated in a previous study
with a sample of a thousand students (Acevedo and Oliva, 1995 ). The TOLT consists of a
ten item paper and pencil test in which the following frameworks of reasoning are
evaluated: proportions, control of variables, probabilities, correlations and combinations.

Interviews

During the second stage of the research, individual interviews were carried out with the
aim of clarifying the sense of the connections found between some of the conceptions. The
proposed questions were similar to those included in the questionnaire and were presented
in an open format on individual sheets. Each interview lasted 15 minutes approximately.

Data analysis

The questionnaire data was processed by means of descriptive techniques and correlational
analysis as well as principal components methods. The statistical treatment was carried out
by means of SPSS programmes. The interviews were recorded and then transcribed
following the habitual procedures for these cases.

Construction of scales

Scales for conceptions in mechanics


The first approach to the study of the underlying alternative frameworks in the answers
consisted of the grouping of items into scales according to content. Subsequent use of
these scales produced more reliable data than that obtained from separate items (Licht and
Thijs, 1990).
A selection of the most representative items was carried out beforehand, in
accordance with the following criteria:

 All the items that included coherent conceptions from a scientific point of view
were selected.
 Items that corresponded to alternative conceptions of an acceptance level that
was not excessively low were also selected. The value of -1 was chosen as the
cut-off point, which in the questionnaire Likert scale represents a partial
disagreement level.
 Items that did not show a significant correlation level with their equivalents in
the other test were excluded.

Thirty two of the 42 items were selected and divided into seven groups according to
thematic affinity, one for each of the tasks involved in the questionnaires. The validity of
the groupings was evaluated later by means of the principal components method. In the
first analysis we considered only one factor as the solution for each analysis. The level of
adjustment to the model was good in every case except for the group of items
corresponding to problem 5 in the questionnaires. In this case, it was necessary to
introduce a second factor to obtain an acceptable adjustment level. Finally, it was possible
to summarise the collected information in only eight factors or scales, with each item
contributing to one and only one of the scales. The composition of each scale, its
reliability and other data of interest are included in table 2.

[Insert table 2 about here]

Almost all the scales were of a bi-polar nature with some items loading positively
on the factor and others, negatively. An exception can be found in scale 5b in which the
only two conceptions correlated positively, together forming one extremity of a continuum
whilst the other opposing extremity remained unrepresented. However, in order to ensure
symmetry with the other scales, this scale was defined as a continuum between two poles:
the low scores for the factor were assigned the opposite significance to that established for
the high scores.
Each of the scales was defined by adding the partial scores of the corresponding
items together. The values were later recodified so that they would fit a +1, -1 range,
which is easier to manipulate and to interpret. The assigning of the positive and negative
poles of each scale was realised by using, as a criterion, the proximity of the extremity to
the scientific point of view. This later served to avoid working with negative correlations
among the variables.
As regards the significance of the scales, seven were relatively easy to interpret.
One extremity that was coherent from a scientific point of view and another opposing
extremity offering an alternative vision were identified (Scales 1, 2, 3, 4, 5a, 5b and 7).
The remaining scale (scale 6) was the most complex and problematic to interpret. It
was hoped that this scale would help to evaluate the position of the student between the
Newtonian vision of the concept of force (C6.2) and the association force-movement of the
Aristotelian-scholastic standpoint (C6.1 and C6.3). However, the data obtained obliged us
to reconsider its significance. Surprisingly, items C6.1 and C6.2 of both questionnaires
correlated positively between each other, becoming integrated in the same extremity of the
scale on whose opposite end the C6.3 items were situated. Due to this, in the final
interpretation it was considered that this dimension would mark the tendency to deal with
the problem analytically, considering the possibility of various forces acting at the same
time as opposed to the tendency to do so with only one force in mind. The first of these
extremities was the one which best correlated with the positive pole of the other scales; for
this reason it was assigned the positive pole.
With regard to the results obtained, it must be pointed out that the distribution of
scores in the different scales approached to the normal curve, with mean values that are
very close to the central value of the scale in some cases (Scales 1 and 3), values slightly
displaced towards one of the poles in others (scale 5b, scale 7) and values that clearly tend
towards one extreme or the other in the rest (scales 4 and 6, and scales 2 and 5a
respectively).
Scales for general reasoning skills

The reliability values of the TOLT general scores and their different subscales are
shown in table 3.

[Insert table 3 about here]

The results are in line with those obtained in a previous study on the validation of
the translation of this test into Spanish (Acevedo and Oliva 1995). However, in this case
the mean values seem to be somewhat lower than those of the latter study.
In comparative terms, the simplest reasoning frameworks correspond to variable
control and combinatory reasoning frameworks. The frameworks of calculation of
proportions, probability and correlations yield the lowest results.

Analysis of the relationships between scales

Table 4 shows Pearson´s correlation matrix for scales for mechanics and TOLT.
The correlations found were generally low but reached statistical significance in a third of
the cases.

[Insert table 4 about here]

When an analysis of principal components is carried out, a solution with four


factors that explain approximately 61% of the total variance is obtained. Table 5 shows
information about the factorial loads of the solution rotated by means of the VARIMAX
method. From the point of view of scales for mechanics, the first factor loaded on scales
5b and 6, the second factor loaded on scales 4 and 7, the third factor loaded on scales 1 and
3 and the fourth factor loaded on scales 2 and 5A. The TOLT appears to be spread
between factors 1 and 2, thereby indicating the scales with which it maintains a significant
relationship.

[Insert table 5 about here]


A more profound study of the data led to an analysis of the relationship between
scales for mechanics 4, 5b, 6 and 7 and the different TOLT subscales (Table 6). The
analysis of principal components yielded a solution with three factors that explained 54%
of the total variance. In this case it seemed advisable to carry out rotation by means of the
OBLIMIN method which does not require factors to be orthotogonal among themselves.

[Insert table 6 about here]

The first factor appears to be related to a general factor of formal thought.


Furthermore, the mechanics scales 5b, 6 and 7 load slightly on it. The greatest weight in
this factor was contributed by the framework of proportional reasoning and by correlation
analysis.
The other two factors refer to more specific aspects of the tasks. Scales 4 and 7
again appear on one side, whilst scales 5b and 6 appear on the other. It can be seen that
these two factors of greater specificity are also weakly influenced by some of the general
reasoning frameworks. Scales 4 and 7 load on the control of variables and combinatory
reasoning, while scales 5b and 7 together load on the combinatory reasoning.
Finally, significant relationships (p<0.01) were detected between factor one and
the other factors, although the detected values in this case were also small. The
correlation was 0.22 between factors one and two and 0.23 between factors one and three.

The results of the interviews

Due to limitations of space, only aspects of the scales for mechanics 4 and 7 shall be
commented on, as their relationship was the most difficult one to interpret. As regards
scale 4, during the interviews a drawing of a ball being whirled in a circular movement
at the end of a string on a vertical plane was presented. The string breaks the moment it
reaches a horizontal position on the upward movement of the ball. The task consisted of
drawing the new trajectory of the ball as from the instant the string breaks. Some students
identified the central action of the string as being a relevant variable of the phenomenon.
These students tended to perceive a change in the behaviour of the system the moment the
string breaks. This led them to conceive a change in the trajectory, which changed from
circular to rectilinear and tangential to the trajectory:

‘...It’ll go off in the direction of the last position, which is a straight line. The string
doesn't make it go round any more’ (Juan).

‘As there is nothing to stop it from doing otherwise, it will go off in a straight line’ (Sara).

‘The string has broken... now there's nothing to make it follow a curved trajectory’ (Pedro).

This contrasts with another group of pupils who drew a curved trajectory that was similar
to that followed by the ball up to that point. These students did not pinpoint the “string
variable” as the key to the phenomenon, nor did they perceive the effects caused by the
breaking of the string:

‘It tends to continue the path it has been following’ (Adela).

‘As the ball is used to going round... it will tend to continue doing the same’ (Carlos).

‘The ball will try to continue its circular movement for some time’ (Eva).

As can be seen, while there are reasons based on natural or intrinsic tendencies of a body
there also are nuances that recall anthropomorphic or animistic thought.
On the other hand, with regard to scale 7, during the interviews students were asked
to predict the time it would take an object to fall from a tenth floor window and from a
fifth floor window. The aim was to find out what ideas would be expressed when they
came to making their predictions. Again, a number of students were capable of perceiving
a change in an essential variable of the phenomenon i.e. the velocity of the object. These
students reached the conclusion that the falling time for an object dropped from the tenth
floor would not be twice that of an object dropped from the fifth floor, thus concluding that
a falling object increases its velocity. When this occurred, reasoning was based on an
analysis of variables such as gravity or the acceleration of the object.
‘...The body accelerates as it was at a standstill and then begins a descending movement’.
(Sara).

‘Its velocity changes constantly due to gravity’ (Carmen)

‘As it descends, it speeds up due to terrestrial attraction’ (Juan)

Among the students there were those who considered the fall to be a uniform movement.
No explicative arguments were offered; answers were reiterated, or basic reasons of the
type “it’s logical” or “it’s normal” were given.

Interviewer: ‘…do you think the time has doubled?’


Alberto: ‘ …(doubtfully) I think so... (silence)’
Interviewer: ‘So that's what you think... what are your reasons?’
Alberto: ‘It's normal, so I think it should happen... bodies tend to fall. Stones try to reach
the ground in the shortest time possible, no matter what height they drop from’.
Interviewer:...
Alberto: ‘It's pure logic. If the height is doubled, the time a body takes to fall is also
doubled’.

Others in the same group admitted that a falling object represents movement with
acceleration, although they had difficulties isolating and identifying the variables involved
in a phenomenon. This caused them to ignore this change and induced them to employ a
simple proportionality framework of the double/half type.

Interviewer:...
Rafael: ‘It’ll take twice as long. It makes no difference whether it’s from the tenth to the
fifth or from the fifth to the ground. The time is always the same’.
Interviewer: ‘Why do you think the time is always the same?’.
Rafael: ‘I don’t know how to explain it...I don’t know. I think it’s the way it must be’.
Interviewer: ‘Do you think the stone always falls at the same speed ?’.
Rafael: ‘No. The stone picks up speed as it drops. It falls faster and faster’.
As can be seen, in tasks related to scales 4 and 7 two opposing tendencies emerge
in the students’ line of argument. On one hand, there are explanations based on logical
causal analysis of the phenomenon, contributing explicit arguments that take into account
the different variables that intervene. On the other hand, there are explanations based on
more or less personal intuition that manifest natural or animist tendencies.

Interpretation of factors

Each of the factors delimited in the first factorial analysis shall now be defined and
interpreted. The analysis of these dimensions was carried out taking into account the
following aspects:

(i) The scales for mechanics which integrated each factor.


(ii) The influence that the TOLT exercised on the corresponding factor.
(iii) The explanations contributed in the interviews.

Dimension 1
Consisting of scales 1 and 3, this dimension appears to correspond to a bipolar framework
of conservation or non-conservation of weight when a variable of the phenomenon is
altered. This is a generic framework contained within another of a more general
conservation type mentioned by Piaget and also taken up by Pozo (Pozo et al 1991) as
being one of the structural references that differentiate scientific knowledge from the
implicit theories of adolescents.

Dimension 2
Made up by scales 2 and 5a, it implies the differentiation or identification of the magnitude
of a cause and its effects. In scale 2, the students were assessed according to the grade of
association they established between the weight of an object and the speed with which
gravity acts. Most of them considered the heavier body to be “more sensitive” to gravity. A
similar framework also appeared in scale 5a, in which the students´ conceptions with
respect to the equality or non-equality of forces in an interaction was assessed. Generally,
students tended to think that the body that suffers the greatest effect in a collision is also
the one a greater force acts upon. Students are more reluctant to accept that the intensities
of the forces that participate in an interaction are equal, independently of the effect they
produce and the weight or velocity of the objects that collide.
This dimension may be linked the rule of reasoning described by Stavy and Tirosh
(1966) as it indicates the tendency students have to give reasons of the type “The more of
A... more of B”. It may also be associated with the model of reasoning of the linear causal
type (Viennot and Kaminski, 1991; Rozier and Viennot, 1991; Pozo et al.1991), as there is
an underlying linear relationship between two variables instead of a more complex
relationship that would admit additional variables.

Dimension 3
Consisting of scales 5b and 6, it corresponds to a framework that distinguishes between the
use of various forces rather than only one in the interpretation of a phenomenon. In the
case of scale 5b, the tendency to conceive forces in a single direction during interaction as
opposed to the tendency to do so in a bi-directional manner may be distinguished.
Meanwhile, scale 6 reflected the inclination of the students to think in terms of a single
force or of various forces to explain the behaviour of a moving object. This dimension is
probably linked to a general framework of interaction opposing another based on
relationships of simple linear causality. In fact Pozo and Gómez-Crespo (1997) have
already pointed out this double aspect of linear causal reasoning.
The relationship between this dimension and formal thought is logical, especially
with combinatorial reasoning. It should be taken into account that the multiple causality
framework must be more exacting in its cognitive demands than the linear causality
framework, in which a single cause precedes a single effect, and only simple covariance
rules are needed for its assimilation. The demands within the multiple causality framework
are more complex, requiring, amongst other things, the combination of all the factors that
may be of influence and the joint estimation of the effect produced by all of them. Thus, it
is logical to assume that the pupils that employ combinatorial type reasoning are more
likely to conceive a system in terms of various forces, as the obtained data seems to
indicate.

Dimension 4
This dimension includes scales 4 and 7, and was the most difficult to interpret. Although at
first sight it seemed to be linked to specific aspects, its correlation with the TOLT results
indicates that this was a case of a framework with a higher degree of generality.
Effectively, its relationship with the control of variables and with combinatorial
reasoning leads one to believe that its variability depends on the capacity of the student to
analyse the relevant variables of a system. Through this dimension each student reveals
his/her tendency to tackle problems concerning movement in a logical manner, by means
of more or less articulated causal arguments, as opposed to the tendency to do so in a
intuitive way without control of the variables that intervene in the phenomenon. In the first
case this would lead them to perceive the relevance of a change in one of the variables of
the phenomenon, such as the main action which made the ball follow a circular movement
or the variation in the velocity it fell at. In the second case, the type of reasoning employed
is quite similar to the causal framework proposed by Taber and Watts (1996) to interpret
what for the students is obvious or natural in physical phenomena, or similar to the
intuitive reasoning discussed by Acevedo (1990) and Touger et al (1995)

Conclusions

The results of this study contrast the fertility of the principles formulated by Monk (1995)
regarding research into students’ conceptions in science. In particular, they show the utility
of the principles of structure, scale, orthogonality and probability as instruments which
serve to widen our knowledge in this type of study. In fact, the theoretical framework and
the methodology used were guided by the following four principles:

 Structure: We have presumed that the different answers offered by the students
are in some way structurally linked to each other.
 Scale: We have combined the results for various items in order to throw some
light on these structures.
 Orthogonality: The principle components method was employed in order to
detect the cognitive dimensions used to ‘weave’ these structures.
 Probability: A certain degree of uncertainty in the students’ thought processes
has been assumed to explain the variations which are usually detected in their
answers when these are examined individually.

The experimental data obtained indicates the existence of a certain degree of


structuralization in the students’ answers. This ‘organization’ is far from being simple and
may configure a hierarchical network in which certain structures fit into others in
accordance with an order that ranges from greater to lesser generality.
We have delimited a set of eight scales which explain some of the results obtained
in the majority of the items of the test. These scales have been defined in a bipolar manner
with one extreme which normally delimits conceptions that are close to a scientific point of
view and the other extreme which delimits alternative conceptions considered unscientific.
Furthermore, we have detected a relationship between pairs of scales and also a
relationship between pairs of scales and the general frameworks of formal thought. This
has been interpreted in terms of reasoning patterns of a certain degree of generality that
would be shared to a greater or lesser degree by certain conceptions of mechanics. Four
factors or dimensions were isolated and configured round the following frameworks:

(i) Conservation/non conservation of weight in different situations (dimension 1)


(ii) Differentiation/comparison of the magnitude of a cause with the magnitude of its
effects (dimension 2)
(iii) Various forces/a single force in the study of the dynamic behaviour of a system
(dimension 3)
(iv) A logical-analytical vision/an intuitive vision of the type of movement of an object
in the study (dimension 4)

The dimensions explored agree with certain causal rules dealt with in other studies with
regard to the explanation of certain similarities that are observed in the conceptions of
different topics. More specifically, the first three seem to be coherent with regard to two of
the structures of thought postulated by Pozo (Pozo et al 1991) distinguishing the
characteristics of the implicit knowledge of common sense. They are in line with the
conservation versus non conservation framework and with the dimension concerning the
interaction versus simple linear causality, respectively.
The fourth dimension also seems to have an important general component and may
form part of a more extensive causal tendency of the type proposed in some of the studies
mentioned (Acevedo, 1990; Touger et al., 1995; Taber and Watts, 1996).
As regards the degree of structuralization of the different conceptions, the results
obtained indicate behaviour between that which is established by the formal operations
approach and that predicted by the more fragmentary approach of the alternative
conceptions (Pozo el al., 1992). However, the degree of generality found in some causal
frameworks may be sufficient to enable one to postulate, within certain limits, the
possibility of changes in certain conceptions once others that share certain features of the
implicit structures are changed. Therefore, overcoming alternative conceptions would
imply more than a local change in specific contents. It would involve a change in the
implicit theories within which the replies of the students underlie and consequently a
change in the mental structures to be found behind them (Pozo el al., 1992). These
conclusions seem to agree with some of the data obtained in research involving this
question (Gómez-Crespo et al.,1995).
Although the acquisition of these structures is essential for a change in thought,
they are not a guarantee of success as the specific conceptual aspects of each domain in
particular are also of importance (Vosniadou, 1994).
A change in knowledge seems to involve a transformation at two levels: one that
implies a large scale change in the causal framework of the individual, and another that
operates by means of a change in the specific contents. This demands greater effort on the
part of the teacher who should also pay attention to these more general frameworks that
underlie many of the replies found on specific issues.
Finally, it is clear that it is necessary to carry out a careful re-examination of
current models of conceptual change, as well as the teaching strategies derived from these
same models.
References

Acevedo, J.A., 1990. Razonamiento causal en una tarea de contexto natural. Un estudio
evolutivo con estudiantes de bachillerato. Investigación en la Escuela, 10, 61-70.
Acevedo, J.A. and Oliva, J.Mª.,1995, Validación y aplicaciones de un test de razonamiento
lógico. Revista de Psicología General y Aplicada, 48, 339-352.
Andersson, B., 1986, The experiential gestalt of causation: a common core to pupils´
preconceptions in Science. European Journal of Science Education, 8(2), 155-171.
Closset, J., 1983, Sequential reasoning in electricity. In Research on Physics Education:
Proceedings of the First International Workshop (Paris: CNRS), 313-319.
De Posada, J.M., 1997, Conceptions of high-school students concerning the internal
structure of metals and their electric conduction: Structure y evolution, Science Education,
81, 445-467.
Driver, R., Guesne, E. and Tiberguien, A.,1985, Children´s ideas in Science (Milton
Keynes: Open University Press).
Engel clough, E., and Diver, R., 1986, A study of consistency in the use of students´
conceptual framework across different task contexts. Science Education, 70(4), 473-493.
Gómez-Crespo, M.A., Pozo, J.I. and Sanz, A., 1995, Students´ ideas on conservation of
matter: effects of expertise and context variables. Science Education, 79(1), 77-93.
Hierrezuelo, J. and Montero, A., 1991. La ciencia de los alumnos. (Vélez-Málaga, Spain:
Elzevir).
Hackling, G.M. and Garnett, P.J., 1985, Misconceptionss of chemical equilibrium.
European Journal of Science Education, 7(2), 205-214.
Jiménez-Aleixandre, M.P., and Fernández, J., 1989, ¿Han sido seleccionados o se han
acostumbrado? Ideas de los estudiantes de biología sobre selección natural y consistencia
entre ellas. Infancia y Aprendizaje, 47, 67-81.
Licht, P. and Thijs, G.D., 1990, Method to trace coherence and consistence of
preconceptions. International Journal of Science Education, 12(4), 403-416.
Maloney, J., 1984, Rule-governed approaches to physics Newton´s Third Law. Physics
Education, 19, 37-42.
Maurines, L., 1992, Spontaneous reasoning on the propagation of visible mechanical
signals. International Journal of Science Education, 14(3), 279-293.
Monk, M., 1995, On the identification of principles in science that might inform research
into students' beliefs about natural phenomena. International Journal of Science
Education, 17(5), 565-573.
Oliva, J.Mª, 1994, Influencia de las variables cognitivas en la construcción de
conocimientos de mecánica. Un estudio empírico y un análisis computacional.
Unpublished PhD thesis, Universidad Nacional de Educación a Distancia. Madrid.
Pozo, J.I. and Gómez Crespo, M.A., 1997, ¿Qué es lo que hace difícil la comprensión de la
ciencia? Algunas explicaciones y propuestas para la enseñanza. En L. del Carmen
(Comp.), La enseñanza y el aprendizaje de las Ciencias de la Naturaleza en la Educación
Secundaria (Madrid: Horsori), 73-105.
Pozo, J.I., Gómez-Crespo, M.A. and Limón, M. y Sanz, A., 1991, Procesos cognitivos en
la comprensión de la Ciencia: las ideas de los adolescentes sobre la Química (Madrid:
CIDE).
Pozo, J.I., Pérez, M.P., Sanz, A., and Limón, M., 1992. Las ideas de los alumnos sobre la
Ciencia como teorías implícitas. Infancia y Aprendizaje, 57, 3-22.
Rozier, S. and Viennot, L., 1991, Students´ reasonings in thermodynamics. International
Journal of Science Education, 13(2), 159-170.
Stavy, R. and Tirosh, D., 1996, Intuitive rules in science and mathematics: the case of
“more of A-more of B”. International Journal of Science Education, 18(6), 653-667.
Taber, K.S. and Watts, M., 1996, The secret life of the chemical bond: students´
anthropomorphic and animistic references to bonding. International Journal of Science
Education, 18(5), 557-568.
Tirosh, D. and Stavy, R., 1996. Intuitive rules in science and mathematics: the case of
“everything can be divided by two”. International Journal of Science Education, 18(6),
669-683.
Tobin, K.G. and Capie, W., 1981, Developmental validation of a group test of logical
thinking. Educational and Psychological Measurement, 4, 413-424.
Touger, J., Dufresne, J., Gerace, W., Hardinan, P. and Mestre, J., 1995, How novice
physics students deal with explanations. International Journal of Science Education, 17(2),
255-269.
Viennot, L., 1985, Analysing students´ reasoning in Science : a pragmatic view of
theoretical problems. European Journal of Science Education, 7(2), 151-162.
Viennot, L. and Kaminski, W., 1991, Participation des maitres aux modes de raisonnement
des élèves. Enseñanza de las Ciencias, 9(1), 3-9.
Viennot, L. and Rainson, S., 1992, Students´ reasoning about superposition of electric
fields. International Journal of Science Education, 14(4), 475-487.
Vosniadou, S., 1994. Capturing and modelling the process of conceptual change. Learning
and Instruction, 4(1), 45-69.
Watts, M. and Taber, K.S., 1996, An explanatory gestalt of eseence: students´ conceptions
of the ‘natural’ in physical phenomena. International Journal of Science Education, 18(8),
939-954.
Table 1. Topics dealt with in the different questions and corresponding conceptions.
Topics implied in the items Conceptions Context
Questionnarie A Questionnarie B
1. Influence of the topology C1.1 Bodies weigh more when they are placed on objects Two balls joined by A ball placed on a spring
of a system on weight. that support them. means of a string, both or hung from it.
C1.2 Bodies weigh more when they are hung from placed on a hand or
objects that support them. hung one on either side
C1.3 The weight of bodies remains invariable of the hand.
irrespective of the manner in which they are supported.
(*)
2. Influence of weight on the C2.1 Lighter objects are more easily accelerated by Two balls are thrown Two balls are dropped
vertical movement of gravity. upwards at the same from the same height.
objects. C2.2 Heavier objects are more easily accelerated by velocity.
gravity.
C2.3 The weight of bodies does not affect acceleration.
(*)
3. The relationship between C3.1 The absence of air does not affect weight or An object dropped in An object on a weighing
atmosphere and gravity- gravity.(*) the interior of a room machine enclosed in a
weight. C3.2 If there is no air, there is no gravity or weight. from which all air has container from which all
C3.3 If there is no air, there is no gravity yet the weight been removed. air has been removed.
remains unaffected.
4. The trajectory of an object C4.1 The object tends to continue its circular movement. A ball is tied to a string A ball is ejected through a
being whirled in a circular C4.2 The object flies off in a straight line, in a non and whirled in a circular shaped tube.
movement at the end of a radial, oblique direction (affected by inertia and circular movement.
string after it is suddenly centrifugal action). The string suddenly
released. C4.3 The object flies off in a straight line tangent to the breaks.
path. (*)
(*) Indicates the theoretically most appropriate option from a scientific point of view.
Table 1. [continuation]

Topics implied in the items Conceptions Context


Questionnarie A Questionnarie B
5. Equality or not of forces in C5.1 Only the heavier body which initiates the action A truck travelling at A boy kicks a football.
interaction (collisions). exercises a force. great speed collides with
C5.2 Both objects exercise mutual force, but the heavier a parked bicycle.
object which initiates the action exercises a greater force.
C5.3 The forces exercised by two bodies are always
equal. (*)
6. Analysis of forces in C6.1 There is only one force which acts in the same A ball is thrown A football in an
movements of ascent, under direction as the movement. vertically upwards. ascending trajectory
the influence of gravity. C6.2 Only one force: the weight. (*) after being kicked.
C6.3 Two forces: the weight and another in the direction
of the movement.
7. Velocity of falling objects C7.1 The object travels equal distances in equal periods A free falling object A ball rolling down an
under the influence of of time. dropped from a certain inclined plane without
gravity. C7.2 The object travels an increasingly greater distance height. friction.
in a shorter period of time.(*)
C7.3 The object travels an increasingly shorter distance
in a longer period of time.
(*) Indicates the theoretically most appropriate option from a scientific point of view.
Table 2. Details of the scales.
Scale Pole (+) Pole (-) Number of Cronbach Mean DS
items  score
[+1,-1]
S1 Conservation of weight in variations of Bodies weigh more when hung.C1.2(A) 4 0.60 0.06 0.45
system configuration. C1.3(A) C1.3(B) C.1.2(B)

S2 Independence of the movement under the The heavier objects are more affected by 4 0.73 -0.31 0.53
influence of gravity with respect to the gravity.C2.2(A) C2.2(B)
weight of the object C2.3(A) C2.3(B)

S3 Independence of weight with respect to the Bodies have no weight in the absence of 4 0.64 0.08 0.46
presence of air or no air.C3.1(A) C3.1(B) air.C3.2(A) C3.2(B)

S4 Rectilinear inertia tangent to the Circular inertia. Bodies that whirl round in 4 0.48 0.26 0.40
trajectory.C4.3(A) C4.3(B) a circle continue moving in the same
trajectory when released. C4.1(A) C 4.1(B).

S5A Equality of forces in a collision. C5.3(A) The object of greater mass and velocity 4 0.71 -0.58 0.35
C5.3(B) exercises greater force in a collision.
C5.2(A) C5.2(B)

S5B In a collision two bodies exercise force on Only the object of greater mass and velocity 2 0.54 -0.13 0.57
each other. exercises force in a collision. C5.1A)
C5.1B)

S6 Various forces act on a body ascending Only one force acts on a body ascending 6 0.70 0.36 0.39
vertically. C6.3(A) C6.3(B) vertically. C6.1(A) C6.1(B) C6.2(A)
C6.2(B)

S7 A free fall movement accelerates. C3.2(A) A free fall movement remains at a constant 4 0.76 0.25 0.53
C3.2(B) velocity. C3.1(A) C3.1(B)

(A) and (B) refer to the questionnaires from which the item has been taken.
Table 3. TOLT data and corresponding subscales.
Scale Number Cronbach Mean DS
of items  score
Proportions (PROP) 2 0,78 0,58 0,82
Control of Variables (CVAR) 2 0,77 0,77 0,88
Probabilities (PROB) 2 0,57 0,26 0,56
Correlations (CORR) 2 0,49 0,52 0,71
Combinatory (COMB) 2 0,59 0,73 0,81

TOLT 10 0,76 2,87 2,48


Table 4. Correlation matrix for the scales on mechanics and the TOLT.
S1 S2 S3 S4 S5A S5B S6 S7 TOLT

S1 1.00 0.05 0.20** 0.09 -0.03 0.06 0.15* 0.06 0.07

S2 1.00 0.06 0.09 0.22** 0.03 -0.01 0.02 0.10

S3 1.00 0.11 -0.04 0.04 0.14 0.02 0.08

S4 1.00 0.00 0.09 0.15* 0.24*** 0.18*

S5A 1.00 0.22** 0.13 -0.06 -0.01

S5B 1.00 0.32*** -0.07 0.30***

S6 1.00 0.09 0.34***

S7 1,00 0.26***

TOLT 1.00
*** p< 0,001 ** P< 0,01 * P< 0,05
Table 5. Principal components analysis for the scales on mechanics and the TOLT.

Variables VARIMAX rotation

F1 F2 F3 F4

S5B 0.78

S6 0.73

TOLT 0.60 0.47

S7 0.80

S4 0.66

S2 0.83

S5A 0.69

S3 0.77

S1 0.74
Only factorial load values greater than 0.30 are included.
Table 6. Principal components analysis for some of the scales on mechanics and
TOLT subscales.

Variables OBLIMIN rotation

F1 F2 F3

PROP 0.73

CORR 0.71

PROB 0.64

CVAR 0.56 0.42

COMB 0.53 0.34 0.35

S4 0.80

S7 0.34 0.64

S5B 0,30 0.77

S6 0,31 0.74

Only factorial load values greater than 0.30 are included.


Appendix: Question number two. Questionnaire A

A wooden ball of 50 grams and a steel ball of 500 grams are both thrown vertically
upwards at the same velocity.
Which of the two balls stops travelling upwards first? Ignoring any effect air friction may
have, one may state that:

a.-The 50 gram ball will stop first as it is more easily affected by the braking influence of
gravity.
-3 -2 -1 0 +1 +2 +3
Total disagreement Indecision Total agreement

b.-The 500 gram ball will stop first as it weighs more.


-3 -2 -1 0 +1 +2 +3
Total disagreement Indecision Total agreement

c.-Both balls stop at the same time. Mass exercises no influence as long as there is no air
friction.
-3 -2 -1 0 +1 +2 +3
Total disagreement Indecision Total agreement
Acknowledgements

I would like to thank the Spanish Ministry of Education and Culture for the financial
support provided through their programme of grants for education research.

View publication stats

You might also like