01-principles.Rmd

# Guiding principles

Carpentries curricula are developed based on the results of research in the science of
teaching and learning. We rely on findings synthesized in the book 
[How Learning Works: Seven Research-Based Principles for Smart Teaching](https://www.wiley.com/en-us/How+Learning+Works%3A+Seven+Research+Based+Principles+for+Smart+Teaching-p-9780470484104) [@ambrose2010learning]. We use this text in our Instructor Trainer training program, so that
Trainers (who teach our Instructor Training courses) can understand why we teach the way we
do, and why our lessons are designed the way they are.

The authors identify seven principles of learning (direct quotation from the book are bolded):

1. **"Students' prior knowledge can help or hinder learning."** -- Identifying
  what the learners know before coming to our workshops help us adjust what we
  teach. One way we do this is through [our pre-workshop surveys](https://carpentries.org/assessment/). We also give the learners
  frequent exercises (or "challenges") throughout our lessons, which help Instructors
  and learners identify and correct misconceptions.
2. **"How students organize knowledge influences how they learn and apply what
  they know."** -- Human working memory is limited, and can only handle about
  five to seven separate items of information at a time. We organize our lessons
  to introduce a few concepts at a time, and then provide challenges to give learners
  an opportunity to practice using these new concepts. This lets learners build connections
  between new concepts and their previous knowledge and transfer this new
  knowledge into their long-term memory, 
  increasing the likelihood that they will be able to use this information successfully
  in new contexts.
3. **"Students' motivation determines, directs, and sustains what they do to
  learn."** -- Our learners come to our workshops already motivated to learn the concepts
  we teach. They realize they need the skills we teach to conduct their
  research more effectively. They have experienced the pain that comes with
  copying and pasting data across spreadsheets, or having to re-do complex
  graphs over and over as new data come in. However, they may also be
  intimidated by how much they have to learn before being proficient
  programmers and data analysts. Two strategies we use to keep our 
  learners motivated
  are: (1) to create a positive learning environment, and (2) to teach the most
  useful skills first. We teach [both of these in our Instructor Training program](https://carpentries.github.io/instructor-training/08-motivation/index.html),
  and discuss how they influence curricular design below.
4. **"To develop mastery, students must acquire component skills, practice
  integrating them, and know when to apply what they have learned."** -- 
  Our lessons use frequent challenges to provide
  opportunities for learners to practice applying their new skills. These challenges
  are designed to incrementally build on each other and integrate previously taught
  and new skills. Careful attention to exercise design helps assure learners will
  be able to transfer the
  skills they acquire in our workshop to their own research.
5. **"Goal-directed practice coupled with targeted feedback enhances the quality
  of students' learning."** -- When Learners try to solve the challenges we include in our
  lessons, they receive direct feedback from the computer - either an
  error message or the expected answer. Error messages are often opaque, and
  do not on their own help learners advance in their learning process, making them
  frustrated and demotivated. Our lessons are designed to be delivered as real-time
  in-person instruction, so that learners get feedback from Instructors and workshop helpers
  that is human-parsable and directed to their level of understanding. Furthermore,
  challenges used in a lesson should only require the skills
  that have already been introduced during the workshop, and should have a limited
  range of possible answers.
6. **"Students' current level of development interacts with the social,
  emotional, and intellectual climate of the course to impact learning."** -- Providing
  a positive learning environment reduces learners' stress
  and helps increase their confidence in their ability to use the skills we teach.
  Creating this positive environment is a responsibility shared among all
  participants: Instructors, helpers, workshop hosts, and learners. 
  Setting expectations by introducing 
  [our Code of Conduct](https://docs.carpentries.org/topic_folders/policies/code-of-conduct.html) 
  at the start of each workshop, and enforcing it, 
  contributes to making the workshop a welcoming space for everyone.
  Other strategies we use to create a positive learning environment [are covered in our
  Instructor Training](https://carpentries.github.io/instructor-training/08-motivation/index.html). 
  Curricular content also plays a major role in creating a positive environment:
  examples chosen should not be alienating, skill level must be appropriate for
  the audience, and the examples and challenges must be directly applicable
  for our learners. For instance, when a learner
  creates a visualization that they can directly apply to their own
  data, it reinforces their motivation and favors a positive learning
  climate.
7. **"To become self-directed learners, students must learn to monitor and adjust
  their approaches to learning."** -- In-person workshops allow Instructors to
  model the thinking process that is needed to address the challenges 
  in our lessons. As an Instructor, being very explicit ("thinking
  aloud") about the steps of the mental model that are involved in identifying
  the functions to use, the values of the arguments they take, and the order in
  which to call these functions to solve an exercise, helps learners to think
  of the questions they need to ask themselves when facing new problems to
  solve. While this type of approach works for any level of complexity in the
  challenges we teach, it works best for the most advanced ones, where several steps
  need to be integrated to come to the solution. Before reaching this level of
  complexity, the challenges can be designed to guide this process, using
  scaffolding. Scaffolding is the process of providing support to a learner while 
  new subjects and concepts are introduced.  Scaffolding assists learners as they 
  progress through increasingly complex code creation by breaking the complex code 
  into smaller, more manageable chunks.  Common practice problems used in instructional 
  scaffolding are Parson's problems, where all the pieces of code to answer the problem 
  are already written but are not in the correct order; and, fill in the blanks. 
  Instructional scaffolding might be one of the most important
  things we use in our workshops. It sets learners on a successful path
  for further self-directed learning. When developing the content of the curriculum, think of the
  kind of thinking process that is needed to successfully address the research
  questions in your field.

Applying these principles effectively requires that they are incorporated into both
**what** is taught (content) and **how** it is taught (delivery). Our 
Instructor Training program focuses on teaching Instructors how to use these principles 
in their teaching. In this handbook, we focus on applying these principles to
curriculum design. Before starting to create lesson content, we highly recommend that you familiarize yourself with our [Instructor Training curriculum](https://carpentries.github.io/instructor-training/). 

## Backward design

[Backward design](https://en.wikipedia.org/wiki/Backward_design) is an instructional 
design model that starts with identifying the desired outcomes of a learning experience,
including core skills and concepts that learners need to acquire. These identified outcomes
are used to develop course content and assessments to measure learners' progress towards
these outcomes. This model was developed by Grant J. Wiggins and Jay McTighe in the late
1990s and is expanded on in their text 
[Understanding by Design](https://www.pearson.com/us/higher-education/program/Wiggins-Understanding-by-Design-Expanded-Second-Edition-2nd-Edition/PGM229455.html). We use backward design in developing our curricula because of its focus on identifying clear, measurable
learning goals and providing assessments aligned with those goals. 

In essense, the backward design process has three stages:

1. Identify the practical skills we aim to teach. 
2. Design challenges to give an opportunity for our learners to practice and
   integrate these skills.
3. Identify what we need to teach for our learners to acquire these skills.
   
This approach ensures that all the skills we teach work together to meet the over-arching
goals of our curriculum. It also reduces the risk that we won't teach a concept learners
need in order to be able
to master the skills we aim to teach. Similarly, it avoids teaching topics that do not help us (and our learners) meet our goals.

Reducing distractions is part of our lesson design as we strive to reduce cognitive load on learners. To this end, we also develop our lessons to be centered around a narrative and a dataset they can relate with quickly.

Because our workshops are domain-specific, the data we use, and the type of questions we ask
with the data are already somewhat familiar to our learners. Their energy and focus can be
directed towards learning the skills we teach rather than on getting familiar with data and
concepts that are foreign to them. This strategy also increases the motivation of our
learners. By learning how to solve problems that are familiar to them, they can more easily
transpose these skills directly to their own data, and have a good starting point to
continue their learning process as they try to solve new or more complex problems with their
own data.

### Identifying the practical skills

Our primary aim in a Carpentries workshop is to increase the confidence of our learners. We
want to demystify and make accessible the process of computing and analyzing data. More than
a third of learners at our workshops have little to no coding experience [@jordan2018assessment]. Our workshops provide them an
opportunity to try, in a friendly environment, something they perceive as intimidating.

Another important goal is to make the research life of our learners easier. We
emphasize teaching "good enough practices" [@wilson2017goodenough] - concrete
skills that are accessible, able to be adopted by researchers of any skill
level, and likely to make an immediate positive impact on learners' work.
Teaching defensive programming, how to use spreadsheets effectively, or how to
organise files consistently across research projects, are practical skills that
can save a lot time when learners apply them in their own research.

When developing a new curriculum, the first step is to identify the skills that will be the
most immediately useful to learners and have the biggest impact on their work. This will 
vary a lot, so having a clear idea of your lesson's intended audience is critical at this
stage.

We will discuss in detail the process of defining your 
audience and identifying these core skills for your lesson
in [a later chapter][Deciding what to teach].

### Designing challenges to assess understanding

Once you have identified these high-impact skills, the lesson content should be designed to
create frequent opportunities for learners to practice these skills while exemplifying the
tasks they perform in their daily work. Live coding and hands-on challenges that learners
can directly relate to should allow them to envision how they can start using the skills
taught with their own data as soon as the workshop is over.

In traditional Western instruction, learners are presented with new material during course time
and then sent home to practice applying the concepts learnt on their own. A major limitation of this
approach is that learners often encounter difficulties in trying to apply their new knowledge or skills
and need to troubleshoot on their own, without support. Education research shows that novices learn best when
they are given feedback and coaching in real time while practicing their new skills (see principle
number 5 above) so that errors are corrected and mis-steps redirected before mistakes have a chance to become
discouraging or engrained in learners' memory.

To this end, Carpentries workshops are designed to provide frequent opportunities for learners to practice new skills. To be
helpful in providing useful feedback, these challenges need to both be:

1. narrowly targeted to the skills that have been taught (i.e. not to depend
   on untaught concepts), and;
2. diagnostic (Instructors should be able to tell what the learner is
   misunderstanding based on how they answer the question).

The practical aspects of creating useful challenge problems is discussed in [a
later chapter][Designing challenges].

### Planning the content of the lesson

After deciding on a list of core skills your learners need, and 
creating a set of well-targeted, diagnostic exercises, you can
then start to create the bulk of the content for the lesson. 
This material can be thought of as the "script" for the 
instructor to follow while teaching and should be planned
very carefully to complement the exercises you've already
designed. We cover the process of creating curricular content
in [a later chapter][Developing content]

## Creating a narrative and selecting a dataset 

Because we strive to provide a realistic experience for learners
that is as similar as possible to the workflow they would use
in their own work, our lessons use real data and are structured
in a natural flow that corresponds to how a learner would experience
working with their data in real life. For many curricula, this
means starting with a lesson on data organisation and progressing
through data cleaning, analysis, and visualisation or reporting. 
It is important to choose a dataset that is an authentic 
representation of what your audience would encounter in their
day-to-day work. The practicalities of chosing an appropriate 
dataset are covered in a [later section][Picking a dataset]
of this handbook.

## Limitations of our approach

Learners can't go from complete novices to experts in two days (or 
in the course of any single class). We aim to provide learners with 
three things:

1. A set of foundational concepts and skills;
1. A mental model that connects those concepts into a useful
framework and that can be built upon in their future learning; and
1. The motivation and skillset they need to continue learning 
past the end of the workshop. 

Managing learners' expectations, and clearly communicating to 
them what they will (and won't) be able to do by the end of the
workshop, is important because it limits the chance of demotivation.