Jupyter Notebook as the Physics Experimental Laboratory’s Logbook

First Approach

Irene Urcelay-Olabarria

, Ruth Lazkoz

, Jon Urrestilla

, Aritz Leonardo

and Josu M. Igartua

Fisika Aplikatua I Saila, Bilboko Ingenieritza Eskola, Euskal Herriko Unibertsitatea (UPV/EHU),

Urkixo Zumarkalea z/g., Bilbo, Spain

Department of Theoretical Physics and History of Science, University of the Basque Country (UPV-EHU),

48040 Bilbao, Spain

Applied Physics II Department, Science and Technology Faculty, University of the Basque Country (UPV/EHU),

B. Sarriena s/n, Leioa (Bizkaia), Spain

Keywords:

Jupyter Notebook, Virtual Laboratory, Python.

Abstract:

In the Physics Degree it is of fundamental importance to practice in an Experimental Laboratory. The standard

Laboratory Sessions consist of two main parts: data handling and data processing. The session should also

have a prologue, where students get to know the underlaying theory of the practical session and an epilogue,

where students present the results obtained and the difﬁculties encountered. The prologue and the epilogue

naturally decouple from the work in the laboratory. Data processing, in most cases, is effectively decoupled

from the work in the laboratory, as well. In this short paper we present a tool, the Jupyter Notebook, an

electronic laboratory logbook, which conveniently facilitates the decoupling of the data handling and process-

ing, but which merges almost completely into an electronic notebook the four parts of the laboratory practical

session: theory, data, processing and presentation. But, interestingly, the notebook goes beyond that: it allows

the students to explore the data in an interactive way (simulating variants), to acquire a deeper knowledge of

the data (by digitally altering the experiment or simulating new ones), to propose new experiments, etcetera.

We strongly believe that this tool can also motivate the students: the results are obtained interactively, imme-

diately, visually, and they can be shared and even improved. Moreover, the laboratory sessions get optimized:

simulations make the sessions be focused on obtaining data and in its variants.

1 INTRODUCTION

1.1 Laboratory in the Physics Degree

In essence, physics is an experimental discipline. In a

very broad sense, it deals with observing Nature to try

to describe it through theoretical models with which

one has to experiment in order to know if they ac-

tually conform to reality. The experimental compo-

nent in Physics is extremely important (Wilcox and

Lewandowski, 2017), it is basic, even for theoreti-

cal physicists. That is why degrees in Physics around

the world have included experimental subjects, more

speciﬁcally, laboratory experiments; in fact, this has

been a subject of thorough study (Bernhard, 2003;

Harms, 2003; Sassi, 2001; Vicentini, 2008; Wilcox

and Lewandowski, 2016; Stanley and Lewandowski,

2016). These laboratory experiments are included

in an ”experimental module” within the degree of

physics, and they share the same general compe-

tences. But there are individual laboratory experiment

subjects in all different years of the degree, since each

set of experiments has associated some speciﬁc com-

petences to them. Common sense dictates that those

speciﬁc competences are to vary so as to follow the

higher degree of skilfulness and maturity the students

display after the leaning process in that very same and

previous academic years.

One of the typical general competences of the ex-

perimental module is to be able to perform a linear

regression with the experimental data obtained in the

laboratory (Hmurcik et al., 1989; Orear, 1991), as

well as to properly interpret the parameters obtained

from the ﬁt. However, many physical quantities do

not vary linearly and, therefore, students have to learn

progressively that linear regression will not always be

helpful. Let us give a speciﬁc example of this: there

are some quantum variables, whose values can be ac-

cessed indirectly in the laboratory, which do not be-

have linearly. Students should then learn that for those

458

Urcelay-Olabarria, I., Lazkoz, R., Urrestilla, J., Leonardo, A. and Igartua, J.

Jupyter Notebook as the Physics Experimental Laborator y’s Logbook - First Approach.

DOI: 10.5220/0006352104580463

In Proceedings of the 9th International Conference on Computer Supported Education (CSEDU 2017) - Volume 1, pages 458-463

ISBN: 978-989-758-239-4

instances it is not a good approach to use linear re-

gression. But, of course, in order for the students to

understand those quantum variables and their context,

they have to have gathered the necessary knowledge

about quantum theory. Moreover, they will also have

to have achieved an appropriate skill level to deftly

use an oscilloscope or a ”Lock-in” (see (Collins et al.,

1974; Ricci et al., 2007; Damyanov et al., 2015) for

a few examples of the use of an oscilloscope for the

acquisition of quantum variables in the undergraduate

physics laboratory).

Another general competence of any standard ex-

perimental module is to be able to graphically show

both the experimental data and their ﬁt via theoreti-

cal models. The speciﬁc competence of the particular

subject of the laboratory experiments of that speciﬁc

course would be to be able to to properly perform the

speciﬁc visualization; for example, to use the correct

variable in the correct context.

1.2 Laboratory in Theory

Laboratory Experiments consist of two fundamental

parts: 2 data collection (second stage) and 3 data pro-

cessing (third stage). It is clear that it should also have

a 1 prologue (ﬁrst stage), the theory on which the ex-

periment is based; and an 4 epilogue (fourth stage),

which includes the presentation of the results obtained

and the difﬁculties encountered. The prologue and the

epilogue are clearly not performed at the laboratory,

are therefore naturally decoupled from the work done

in the laboratory. See ﬁgure 1.

All stages have their own general competences as-

sociated to them: knowledge of the theory, capac-

ity to collect data, ability to process them, ability to

communicate the results obtained. But also speciﬁc

competences according to the level (academic year)

in which the experiment is performed: the theory gets

more involved; the data collection and its treatment

gets more complicated; the presentation gets more so-

phisticated, because, for example, the audience’s sci-

entiﬁc level has increased. This development hap-

pens as the students progress in the degree. How-

ever, not all types of competences have to be acquired

in the laboratory. Normally, those associated with

the prologue and the epilogue are acquired outside:

that is why they are said to be naturally decoupled

from those associated with data collection and data

processing. Of these last two (data collection and

data processing), the only one that must necessarily

be carried out in the laboratory is the data collection.

There is a variant of the traditional experiment for-

mat in which the data are not acquired in the labora-

tory, but instead, data are supplied to the students. In

this format, the students do not acquire competences

related to data collection, but the rest of the compe-

tences are identical to the traditional procedure. This

format is validly used in certain circumstances: in dis-

tance learning courses, for example. Nevertheless,

we should note that in the classical format in which

the students collect the data, the competences involv-

ing data collection are given the status of importance

they deserve since they are related to the gathering of

expertise in routinary laboratory work and to the ac-

quisition of dexterity in the handling of experimental

equipment. In fact, much of the time of the lab ses-

sion is spent collecting data.Hence, why not turn the

lab session into a mere data acquisition session? In

this way the students could be convinced of the great

importance of this stage: Data collection. In“real life”

the acquisition of experimental data is an art: it re-

quires a great knowledge of what is being done (theo-

retical knowledge on which the experiment is based;

theoretical knowledge in which the measuring instru-

ment is based; skills on how the instrument is used;

the different working modes; the data it provides and

if data have to be treated previously to do any other

type of calculations; etc.); it requires detailed atten-

tion to it and, usually, a lot of patience and time.

We would not want to lose sight of the fact that

the data collection includes, from our point of view

at least, data visualization. Collecting data, without

seeing them, without showing them, does not make

sense. Even more, data visualization must ideally be

done at the same time, as they are being collected.

In this way it will be possible to realize if the exper-

iment makes sense, if errors are being made, if the

instrument being used is calibrated, if it works cor-

rectly, if the data obtained are coherent, if they follow

the expected trend or if they are in accordance with

the theory presented in the prologue.

If every time data are collected they are displayed

and they behave as expected, it naturally produces sat-

isfaction. On the contrary, when unexpected behav-

ior is faced, the natural reaction is to think about the

possible reasons why the behavior does not ﬁt the ex-

pected trend: this is the interesting point, this reaction

is the one that sets in motion all the acquired com-

petences and causes others to be acquired (Dounas-

Frazer and Lewandowski, 2016). The anomalous be-

havior may ﬁnd its roots, for instance, in randomness,

deﬁcient measurement, poor equipment adjustment,

bad calibration, misunderstandings, the experiment it-

self could be poorly designed... Then it is natural to

repeat the acquisition of that datum or series of data,

after having changed the conditions in a way dictated

by prior reﬂection, to re-visualize the new data and to

repeat the process of reﬂection if necessary. This pro-

Jupyter Notebook as the Physics Experimental Laboratory’s Logbook - First Approach

459

cess in successive approximations based on the mea-

surement, visualization, correction of the necessary

parameters and, if necessary, repetition of the mea-

surement, is what leads to the correct realization of

the experiment. On the one hand, the processing of

the data is going to produce a totally satisfactory re-

sults. On the other hand, and undoubtedly also, it will

allow the students to actually acquire the associated

competences in a natural way.

It is interesting to compare the procedure with the

action of only producing a table containing just the

necessary information to do some calculations and to

obtain the requested values. Clearly, such a table con-

tributes nothing to the competences of the student. If

anything, it creates boredom and apathy about the lab

sessions. Apathy that increases as the students take

part in more and more lab sessions, which in turn

causes the competences associated with the experi-

ments not to be acquired and, also, causes “going to

the laboratory” to be just an easy and necessary for-

mality: it is enough to show up and collect the data in

order to pass the subject; taking data turns the process

into a mere data collection process, without any other

consideration or reﬂection on their quality, meaning

or usefulness.

Data processing, is traditionally carried out in the

laboratory. In fact, in the ﬁrst year of the Physics de-

gree in our university, in the subject called Experi-

mental Techniques I, the students must perform the

data processing in the same session in which the data

are collected; and the session report has to be sub-

mitted before leaving the lab. Data processing (2) is

very important, since it leads to the cognitive develop-

ment of a large number of concepts (error estimation,

propagation of errors, importance of measuring some

variables instead of others, possible design of experi-

ments ...) and it helps acquire or reﬁne a large number

of skills (differentiation, integration, qualitative cal-

culations, approximations, graphical representations

...). These competences are very useful not only for

lab sessions, but also for the rest of the subjects in the

degree. The data processing stage consists, in turn,

of two sub-stages; namely: 1) the knowledge and the

use of the tools necessary to carry out the treatment,

and 2) the treatment of the actual data themselves.

One must know whether a linear regression must be

made and how it is done, and how the results of the

ﬁtting must be presented: with its errors, which have

been propagated or not, with their signiﬁcant ﬁgures;

making the necessary tables, with the variables indi-

cated, accompanied by their units; constructing the

appropriate graphical representations, with the indi-

cated variables, accompanied by their errors, in the

appropriate scale and with the indicated and adequate

precision of the experiment. It is necessary to know

how the ﬁnal results are interpreted: if the experiment

makes sense, if the variable measured behaves as ex-

pected; if the statistics is adequate; if it is necessary to

repeat some measurement or series of measurements,

and why; if there are several trends, what happens if

some points or series of points are removed, how the

behavior changes and why, whether or not to remove

them; what would happen if it were measured in an-

other interval, or in other measurement conditions (if

possible).

The ﬁrst sub-stage, 3.1, consists itself of another

two levels. First, the knowledge of how to carry out

the task: how a linear regression is done; how the

propagated errors are calculated; how and why the

signiﬁcant ﬁgures are assigned and what they mean;

how the representation window, the appropriate scale

and precision are chosen; which is the dependent vari-

able and which is independent. Second, the action

level: calculations have to be done to obtain the inter-

cept and slope, and their corresponding errors, results

have to be truncated and graphical representations fol-

lowing the criteria dictated by the knowledge of the

ﬁrst level have to be performed. Nowadays the sec-

ond level has typically an automatic character (and

our university is no exception), in the sense that al-

most all calculations are made in a fast, reproducible

and reliable way using computers and more or less

sophisticated computer packages. It is important to

do this in this way because, among other things, the

students are being trained for the real world in which

people do not work with graph-paper to make graphi-

cal representations, nor are the calculations or regres-

sions done by hand.

Obviously this second sub-stage has much to do

not only with the data treatment itself, but with their

visualization, analysis and interpretation, or, in other

words, with the (previous) data collection stage of the

session. As mentioned above, in the second sub-stage,

in the data collection stage, the data obtained are in-

terpreted and decisions about the experiment itself are

made: whether to continue, to stop, to repeat, or to

consider other venues such as other measuring inter-

val... In short, the experiment is being completed, al-

ternatives are being considered, others are being dis-

carded and, some others, simulated. All this allows

students not only to actually learn in-situ, with the

possibility of rectifying, but to do it absolutely mo-

tivated, since they realize that the experiment devel-

ops, either positively (because one gets what was ex-

pected) or negatively (when one knowns what and

why something does not work), all of which allow

competences to be acquired more naturally and eas-

ily.

CSEDU 2017 - 9th International Conference on Computer Supported Education

460

1.3 Laboratory in Practice

Our experience shows that laboratory experiments

have always been considered by the students as “easy

subjects”; subjects that one passes easily and that with

little effort provide a good grade. Students do not take

into account the potential of the subject, all the bene-

ﬁts they can get out of it. Even lately, where consider-

able effort has been put into stressing its beneﬁts and

increasing its difﬁculty, the same thinking continues.

Students of latter academic years are not exception,

they pass the subject without realizing its importance

and without noticing all that can be learned in a lab-

oratory, in general, and with regards to the speciﬁc

concepts that are being developed in the laboratory,

in particular

The manuals for the experiments to be performed

in Experimental Techniques III are very thorough,

clear and concise. They present a theoretical intro-

duction, in which the session is contextualised and

its theoretical basis is explained, the speciﬁc objec-

tives of the session and the material to be used. The

manuals also describe the steps to be followed, and

the tables and graphical representations necessary to

obtain the objectives requested. The manuals are self-

contained. However, instead of using the (in our view)

wonderful tool that the manual actually is, students

use it as a recipe to know what and how to measure

and what and how to represent it, without giving it

any thought, without understanding the reason behind

each step, in order to provide a result fast, and get a

grade. The lack of commitment in the students is ap-

parent: they copy the measured data in random sheets

of paper (that get lost more often than it would be

desirable), leave the lab in a rush with the faintest ex-

cuse, have no shame in leaving just a member of the

group performing the experiment since they will all

share the data.In a nutshell: the experiments are never

performed right, never with the necessary depth, and

the students hardly ever acquire any of the relevant

skills or competences.

Team-work is important, and, arguably, even more

in a laboratory. It is one of the competences associ-

ated to the experimental module. Actually, in all the

laboratory sessions, we aim at having students work-

ing in pairs. However, this strategy makes the atti-

tude of the students in the laboratory inappropriate.

In many cases, only one person in the couple actually

works and, in turn, is the only one that goes through a

learning process. In most cases the work is split into

team members, and it ceases to be a team effort. In

summary, this working system does not achieve pro-

moting team-work, and does not ensure that all mem-

bers of a team acquire the associated competences.

This is aggravated, since the division of labor leads

to specialization (which in some context could even

be interesting): the same students are in charge of the

same type of task in every single session, there is no

rotation of tasks (which, if happened, would help stu-

dents acquire all different competences). It is even

worse, since the specialization is inherited to other

laboratory subjects, of the same and other academic

years, precisely for the same reason: (misunderstood)

efﬁciency. Students seek to work as quickly as possi-

ble, ﬁnish the task and get the grade, without having

the time or the willingness to integrate the knowledge

that each session aims to provide.

All the above arguments prove that there is a dan-

gerous difference between what the lab sessions were

designed and implemented for in the curriculum, and

what they actually achieve: laboratory experiments

do not meet any of the objectives, neither general nor

particular; they do not succeed in providing many of

the competences associated to the module of Exper-

imental Techniques, nor those associated to the spe-

ciﬁc subject in question.

2 PROPOSAL

Therefore, we believe that it is very important that

the 3.1 (Knowledge phase in Data Processing) is ex-

ecuted outside the laboratory and that 3.2 (Treatments

in Data Processing) is integrated into the data collec-

tion stage. The Data Collection stage would then be

formed of the following substages: pure data acqui-

sition/collection (2.1), visualization (2.2), being inte-

grated as a (3.2).

This new re-arrangement highlights that acquir-

ing experimental competences involves taking data

in the laboratory as well as visualizing the data as

they are being taken. In any case, speciﬁc com-

petences related to the theory part of the speciﬁc

session can be acquired “outside” the laboratory, as

well as speciﬁc skills related to the communication

of the results. One of the sub-stages of the data

processing, namely 3.1, can also be executed out-

side the laboratory. In fact, performing those tasks

outside the lab turns the acquired competences into

more general ones, making them more versatile and,

probably, more useful. The other sub-stage of the

data treatment (3.2) is integrated into the data col-

lection (2). Data treatment is almost always decou-

pled from the work in the laboratory. In this pro-

posal we present a tool, the electronic lab notebook

(Jupyter Notebook(Jupyter, 2017)), which naturally

decouples the processing of data (pure data process-

ing, what has been called sub-stage I of this stage)

from pure data collection (collected only without vi-

Jupyter Notebook as the Physics Experimental Laboratory’s Logbook - First Approach

461

Prologue EpilogueData Collection Data Processing

Collect Visualize Knowledge Treatment Preparation Presentation

Collect Visualize

Knowledge

Treatment Preparation Presentation

Knowledge Collect VisualizeTreatment PresentationPrologue

stage 1

stage 2 stage 3 stage 4

Proposal: Jupyter Notebook

Outside the laboratory In the laboratory

sub-stage I sub-stage II

sub-stage I sub-stage II sub-stage II

sub-stage I

sub-stage I sub-stage II sub-stage III

Figure 1: Scheme for the Stages of Laboratory Experi-

ments. First and second rows, classical distribution of

stages and sub-stages (in parenthesis): 1-prologue , 2-data

collection (2.1-collect, 2.2-visualize), 3-data treatment (3.1-

knowledge, 3.2-treatment), 4-epilogue (4.1-preparation,

4.2-presentation). Second row, the sub-stages of stage 1

and 2 have clearly separated. Third row rearrangement of

the second row, just for establishing the proposal. Finally,

fourth row our proposal: merging the knowledge of data

treatment (4.1) with prologue (1) and this new section can

be achieved outside the laboratory. The laboratory work

consists on a bucle: the old 2.1-collection, 3.2-treatment

and 2.2-visualization of data get together into collection-

treatment-visualization. Finally, outside the laboratory, or

not, the presentation of the experiment takes place.

sualization), but at the same time integrates in an elec-

tronic notebook (ENB), almost totally, the four stages

of an experiment: theory, data (collection, visualiza-

tion), processing (calculations/representation and

re-calculations re-representation) and presentation.

Furthermore, it allows concepts to be explored in an

interactive way (simulating variants) deepening the

knowledge about those concepts (digitally altering

the experiment or simulating new ones), proposing

new experiments... The results are achieved interac-

tively, immediately, visually, and can be shared and

improved. In addition, the work in the laboratory is

optimized: it is focused on data collection and its vari-

ants.

We would like to highlight the improvements brought

into the actual realization of laboratory sessions by

the use of a Jupyter Notebook as a laboratory note-

book.

1. It is a real notebook: can be saved, shared, edited

simultaneously by the members of the team, re-

viewed, corrected and evaluated digitally.

2. The manual of the experiment becomes the note-

book of each member.Each member has its own

notebook: revisable, correctable and evaluable in-

dependently (more precise in the assessment pro-

cess).

3. There will be more precision in the acquisition of

the competences, as well, and work will be inde-

pendent and autonomous. Each partner has the

opportunity to make their own simulations, and

then discuss them to ﬁnally admit or discard them.

Each one can record in their notes, in a contin-

uum or in turns, the difﬁculties they have encoun-

tered (individually or as a team) and how to rem-

edy them.

4. The data are automatically checked by the inter-

face itself, thus maximizing a good use of time.

Calculations are executable in each moment; par-

tial data can be saved; it can be re-run with dif-

ferent input data and the results can be compared,

reproducibly and reliably, before the decision to

move to the next stage is taken.

5. The graphic representations are automated, and

can always be modiﬁed according to the user’s

preferences, exported in many formats and saved,

shared and inserted in any technical document ...

We propose two levels of implementation: level 1,

for students enrolled in the Physics degree, and level

0, for all other students. The level 0 is subdivided into

two sub-levels, depending on the studies undertaken

by the students: one sub-level for students of the De-

gree in Chemistry (Faculty of Science and Technol-

ogy) and another sub-level for students of Industrial

Engineering (School of Engineering).

Both, the concept and the goal, of an ENB on the

two levels considered is the same. We have estab-

lished these two levels because we want to measure

the impact on physics laboratory sessions, indepen-

dent of the degree the students are enrolled in, and

which Center they are studying at. We want to cre-

ate a tool for general use in laboratory sessions, with

small variants, depending on the a priori attitude of

the students towards learning Physics, and on their the

computational competences. At level 0, the students

will collect data, enter them in the ENB and the note-

book will be responsible for depicting the data (see

(Eshach and Kukliansky, 2016) for insight on difﬁcul-

ties students encounter on this type of analysis). The

elements of interaction with the visual data needed to

do the analysis properly will be provided: simulat-

ing experiments, seeing the effect of removing points,

exploring how settings change etc. At level 1, the in-

troduction of the data and the analysis of the data will

be different concerning the students’ computing com-

petences. The difference is based on the fact that the

students of the Degree of Physics attend a course in

which they acquire the necessary competences to be

able to interact in a conscious and efﬁcient way with

the ENB. Thus, they are forced/expected to use the

CSEDU 2017 - 9th International Conference on Computer Supported Education

462

skills they have acquired in the previous course by re-

ﬁning the NB themselves.

3 CONCLUSIONS

The using of Jupyter Notebook as the lab-log-

notebook, facilitates the decoupling of the data han-

dling and processing and merges almost completely

into it the four parts of the laboratory practical ses-

sion: theory, data, processing and presentation. Re-

sults are obtained interactively, immediately, visually,

and they can be shared and even improved. The lab-

oratory sessions get optimized: simulations make the

sessions be focused on obtaining data and in its vari-

ants. Moreover, the notebook goes beyond that: it al-

lows the students to explore the data in an interactive

way (simulating variants), to acquire a deeper knowl-

edge of the data (by digitally altering the experiment

or simulating new ones), to propose new experiments.

In our view, the new re-arrangement implied by

the proposal highlights that acquiring experimental

competences involves taking data in the laboratory as

well as visualizing the data as they are being taken;

the speciﬁc competences related to the theory part of

the speciﬁc session can be acquired “outside” the lab-

oratory, as well as speciﬁc skills related to the com-

munication of the results; as part of data processing

executed outside the laboratory, which turns the ac-

quired competences into more general ones, making

them more versatile and more useful.

REFERENCES

Bernhard, J. (2003). Physics learning and microcomputer

based laboratory (mbl): Learning effects of using mbl

as a technological and as a cognitive tool, science ed-

ucation research in the knowledge based society. Dor-

drecht: Kluwer, 313-321.

Collins, L. A., Morrison, M. A., and Donoho, P. L.

(1974). Advanced undergraduate-laboratory experi-

ment on electron spin resonance in single-crystal ruby.

American Journal of Physics 42, 560 (1974).

Damyanov, D. S., Pavlova, I. N., Ilieva, S. I., Gourev,

V. N., Yordanov, V. G., and Mishonov, T. M. (2015).

Planck’s constant measurement by landauer quanti-

zation for student laboratories. European Journal of

Physics, Volume 36, Issue 5, article id. 055047.

Dounas-Frazer, D. and Lewandowski, H. (2016). Nothing

works the ﬁrst time: An expert experimental physics

epistemology. ArXiv e-prints.

Eshach, H. and Kukliansky, I. (2016). Developing of an in-

strument for assessing students? data analysis skills in

the undergraduate physics laboratory. Canadian Jour-

nal of Physics, 94(11):1205–1215.

Harms, U. (2003). Virtual and remote labs in physics educa-

tion. Proceedings of the Second European Conference

on Physics Teaching in Engineering Education, 14–17

June (Budapest).

Hmurcik, L., Slacik, A., Miller, H., and Samoncik, S.

(1989). Linear regresion analysis in a ﬁrst physics lab.

Journal of Physics 57, 135 (1989).

Jupyter (2017). The jupyter notebook. http://jupyter-

notebook.readthedocs.io/en/latest/notebook.html.

Orear, J. (1991). Comment on “linear regression analysis

in a ﬁrst physics lab,” by l. hmurcik et al. American

Journal of Physics 59(1):87-88.

Ricci, M. M., Mazzaferri, J., Bragas, A., and Martinez, O.

(2007). Photon counting statistics using a digital os-

cilloscope. Am. J. Phys. 75(8), pp. 707-712.

Sassi, E. (2001). Lab-work in physics education and in-

formatic tools: advantages and problems. pinto, r. &

surinach, s. (eds), physics teacher education beyond,

57– 64. elsevier, paris. Pinto, R. & Surinach, S. (Eds),

Physics Teacher Education Beyond, 57– 64. Elsevier,

Paris.

Stanley, J. T. and Lewandowski, H. J. (2016). Recom-

mendations for the use of notebooks in upper-division

physics lab courses. ArXiv e-prints.

Vicentini, E. S. E. . M. (2008). Aims and strategies of lab-

oratory work. M. Vicentini, & E. Sassi, Connecting

Research in Physics Education with Teacher Educa-

tion (pp. 1-12). International Commision on Physics

Education.

Wilcox, B. R. and Lewandowski, H. J. (2016). Open-

ended versus guided laboratory activities:impact on

students’ beliefs about experimental physics. Phys.

Rev. Phys. Educ. Res., 12:020132.

Wilcox, B. R. and Lewandowski, H. J. (2017). Students’

views about the nature of experimental physics (arti-

cle). under review Phys. Rev. X.

Jupyter Notebook as the Physics Experimental Laboratory’s Logbook - First Approach

463