First draft
Ross Woods, 2021
This book is about the principles of writing something that was once called “programmed instruction.” It is a kind of educational material that actuallly teaches. In that aspect, it is not like a book, which is passive; it depends on you to pick it up and read it. It was a breakthrough because for the first time, the materials could actually teach in that they mimiced the way a human instructor might teach.
Susan Meyer Markle was an American academic in the field of educational psychology, mostly active in the 1960s. Among other books, she wrote the original best textbook on how to write programmed instruction, and the book itself was in programmed instruction. Its title was Good Frames and Bad.
But things have changed. Back then, programmed instruction used paper pages or clunky machines. Computers are now the best way to provide this kind of instruction. From about 2012, the Massive Online Open Course (MOOC) movement started, and it’s now possible to have 100,000 students all together in a course, which is almost to say that an institution can now train potentially infinite numbers of students. Besides, lessons don't have to be put on the pages of a book; they can use graphics, sound, video or anything else that you can put on a computer screen. Some computer games act as simulations and use very similar principles.
One of the added aspects of internet-based instruction was it could use videos of human instructors. When done well, this creates a more personal interaction, especially if students have an opportunity to actually meet the person sometime.
The most well-known modern versions are the clickthrough video option in YouTube, the lesson module in Moodle, and the EDx MOOCs. But I don't think anybody calls it “programmed instruction” anymore.
This kind of instruction is best for very large numbers of students because lessons can be fully automated.
This kind of instruction is best for theoretical topics, even if it’s the theory of how to do something but not the actual practice. It works well for many levels of higher education, and for the theoretical parts of vocational and professional education. It works where knowing how to do something is adequate without actually having to do it. It doesn't work well for the hands-on component. You could teach over the internet about different kinds of hammers and their uses, but you can't check how well students can swing a real hammer to hit a real nail.
One of the limitations of this kind of programming is that it does not differentiate between recognizing a correct action and actually doing the correct action. For example, the objective is for the student to be able to create and edit a computer graphic. The program might show several videos of persons creating and editing computer graphics. Only one of them is correct, and the others will result in some kind of error. The task requires the student to select the correct one. In other words, it the student selects the correct answer, he knows what to do and how to do it, but has not yet actually created and edited a graphic of his own.
It works for topics that won’t change much due to ongoing research. You wouldn't want to waste so much time and energy writing good instruction if researchers could change the knowledge of the field at any moment. (Writers of some kinds of textbook have the same problem.)
It uses questions or tasks where answers can be predetermined and either correct or incorrect. Consequently, it doesn’t work well for some of the kinds of tasks that instructors need to give students, that is, anything where students can come up with creative and original answers, because computers can’t assess “free expression” answers.
First, planners must define what students must be able to do, so that assessors can know whether students have learnt it. (Just thinking isn't enough.)
Second, student must respond actively, they is, they must do something.
Third, when students respond correctly, they need to be told that they were correct. It's called “reinforcement” because it confirms their response and helps them remember it. It's important because the student is learning something new and might be unsure whether they were correct or incorrect.
Fourth, students must be corrected for incorrect responses, so they don't think that they were right and “learn” their incorrect response.
Fifth, the reinforcement or correction is more effective if it is unambiguous and given as soon as possible after the student responded.
Behaviorism is, in its purest form, the idea that learning is the acquisition of observable behaviors. In many cases, it depends on the idea that information should be divided into small separate facts, each of which can be learnt individually.
Learning depends on conditioning, that is:
One of the main criticisms is that it does not explain what happens in the mind; it only speaks of observable behavior. It also doesn't explain some kinds of learning or how people can adapt their behavior to new situations. In teaching, it often chops knowledge into such small pieces that it loses the big picture.
In its less extreme versions, it is the idea that the goals of learning can be written down unambiguously and used as an objective standard for assessment.
A behaviorist approach may be most useful when:
Teaching strategies:
1. Use forward planners.
A forward planner is a message to students about what to expect in a course or unit. It give a mental map of what to expect and allays any fears of the unknown that they might have.
2. It's not the same as teaching face-to-face.
Creating any kind of online education is not the same as teaching normal face-to-face lessons in a classroom. Writers can’t presume that they can simply move their normal classroom materials online. Creating online materials, and especially “programmed instruction” requires more analysis and planning.
3. If you must include unnecessary material, highlight the part that students must notice to get the point.
Students learn better when their attention is drawn to essential information. This draws the students’ attention away from extraneous material so that they are more likely to think only about the essential elements, and prevent overload.
4. Don’t provide the same information twice.
Any verbal or pictorial information is redundant if it presents nothing new for students; they already have that information. Redundant information makes students think harder without helping them to learn, so it is just extra load.
5. Leave out information that is irrelevant to what students are supposed to learn.
Students learn better when extra material (such as interesting but irrelevant words, pictures, sounds, or music) is excluded, simply because they have to process less information.
6. Split information into bite-size chunks.
Students learn better when complex materials are split into smaller “chunks,” each of which is an item of knowledge. (In the older programmed instruction literature, each item was called a “rule.”) Students can learn them one at a time and construct an adequate mental model without getting overloaded.
Depending on your course, each rule or chunk might be simple or quite complex, but it needs the following characteristics:
Notes
All known programmed instruction (and most other teaching methods) is based on sets of learning objectives,
sometimes called instructional objectives.
These are statements of what the students must be able to do at the end of a learning activity or course. They must:
Learning objectives must be expressed in a way that is useful for assessing whether the student has achieved the objective, so they must be clear and unambiguous. They can refer to students’ thought processes (uderstanding, applying, etc.), but if they do, they must still state what students will do to indicate those thought processes.
The idea of SMART is also useful:
Learning objectives are not intuitive for many people because they describe the destination, not the route to get there. Some teachers are more interested about what they (the teachers) will tell students or the content they will cover. Others are interested in the way it will be taught or how students will study it.
So much has been written on the topic of learning objectives that it is better to consult other existing resources.
Bloom’s taxonomy
is a way of classifying educational objectives according to the kind of thinking they require of students. Ben Bloom published several versions and others have also issued other versions. This version includes several adaptations from the original.
The lower levels are recall, comprehension (with three sublevels) and application.
chunkof information in a format that is easy for them to understand.
What would happen if . . . ?
What else could happen?
What kind of results would you expect?
Who else would be affected by … ?
What kind of responses might you get?
A sequence is the order in which students progress through the material in a lesson or course. Sequencing is the skill of putting items of material in an order that makes them very easy for students to understand. As a course writer, you need to sequence material so that students will see a simple, natural progression in the lessons.
All good sequences have the following characteristics:
It's easy to teach people something they already know. The challenge is to teach them something that they don't know yet. If students think That seems easy.
and That makes sense.
or I can do it now.
then the sequence is probably working. Get a sequence a little bit wrong and students will ask questions to cover your gaps. Get it badly wrong, and your students will be frustrated and confused, and you will probably have to teach the whole lesson again. If that's not enough, you'll find that it is more difficult to correct confused students than to teach it well the first time. Like most of teaching, it is about accurately anticipating how students will respond.
Sequencing applies at different levels:
(By the way, sequencing also applies to "whole of curriculum" design and to articulation between schools, e.g. from elementary school to middle school.)
The number of possible effective sequences is almost limitless, just as there is no limit to the number of sequences that don't work.
Sequences can be idiosyncratic, reflecting the creativity and the personal viewpoint of an individual teacher at a particular time. Some creative sequences might be fun for everybody and make the lessons very easy to teach. But it's not always that simple:
Both deductive and inductive teaching approaches presume that learning comprises two parts:
In the deductive approach, the principle comes first. The course writer provides the principle and then uses a series of examples to explain how to apply it. For both course writer and students, this is fairly easy to do, because you simply tell students the principle.
In the inductive approach, the principle comes last. The course writer poses a question then compares a series of examples to draw a conclusion about a principle. This is much more difficult for both course writers and students. You have to guide students to discover the principle, and students have to figure it out themselves.
In lower level courses, the deductive approach is best. In academically higher level courses, both are very useful.
1. Give students both text and pictures together.
Students learn better from text with pictures than from text alone or from pictures alone.
2. Put the text and pictures close together.
Students learn more effectively when the text and pictures of the same concept are placed close to to each other. They have more difficulty understanding the text or the pictures if they are separated from each other by space on the page because the space requires the student’s brain to work harder to process the information, so it is more likely to overload.
3. Give students the verbal and pictorial information at the same time.
Students learn more effectively when the verbal and pictorial information are given simultaneously. Like the spacial distance principle, students have more difficulty understanding the text or the picture if they are given at different times. The time difference makes the students’ brain work harder to process the information, so it is more likely to overload.
4. Put animation and narration together.
Students learn better from animation and narrated text together than from animation alone or from written text alone. Using two channels expands the working memory and make it less likely that students will overload.
“Cognitive load theory” is a set of principles for designing multimedia. The principles work best for students who are first given new information. Later on, they might know enough about the topic that they can figure it out themselves without much trouble. (Based mainly Sweller, 2005 and Mayer 1997.)
The idea of “cognitive load” is that if materials are easy to understand, that is, students work within their thought and memory limits, students are less likely to get confused and more likely to succeed.
Like many general teaching principles, the above principles depend on “cognitive load theory.” When applied, the “cognitive load” aspect makes learning easier.
A frame is a single learning activity. A basic frame has the following parts:
Select one of the following:
An example of a very simple frame:
Letters can represent numbers. For example, the letter a
could represent 11, so instead of writing 11 you could write a
. The letter b
could represent 3
, so you could write b
instead of writing 3
. Then you could use this to add two numbers together.
If a = 11 and b = 3, then a + b = 14. (That's because 11 + 3 = 14.)
Try this:
If x = 3 and y = 4, what is x + y?
5
6
7
8
If the student selects the correct answer, a message appears:
✔
Correct.
If x = 3 and y = 4, then x + y is the same as 3 + 4, so the answer is 7.
If the student selects the incorrect answer, the following message might appear:
✗
Sorry, not correct.
Remember that x = 3 and y = 4.
In software for learning mode, the page might have some extra enhancements:
In test mode, the page might be as follows:
Frames often ask students to differentiate between examples and non-examples. For instance:
Which of the following are continents?
South America
Hawaii
Asia
Africa
The answer is South America, Asia and Africa. Hawaii is not a continent so it is is a non-example.
Frames with multiple choice responses are often preferred for sophisticated thought forms because so many resesarchers have used them for a long time in large scale summative assessments where students' futures are at stake. They all give students a question or task, and require them to select the correct answer from one of a series that is in random order. Questions can be in text, sound, graphic, video, or a combination. The difference between correct and incorrect answers must be indisputable, because the software will deem correct or incorrect based on predetermined answers.
The presentation of options should not normally contain cues. All answers should be about the same length, so that lengths don't make one appear correct. (If only one is long, it tends to look like the correct one.) Answers must be unambiguous to students who are reading them with no hints from outside the text. This makes many questions quite challenging to write, and they normally require field trials. Students often don't interpret them as the writer intended.
When using multiple choice, it is normal to one correct answer and three decoy answers, although some have four decoys. Decoy answers
are incorrect answers that students may select. The combinations of answers might be:
Choosing one out of four or five answers is not the only possibility. It would not be difficult to have a frame with six options to choose from, and multiple answers are correct. For example, the instructions to students might say, In the list below, select all correct answers.
.
In a learning situation (not a final test), the answer can have up to five decoys and you can give students feedback that is individualized to each answer. That is, if they got it wrong, you can tell them why, and if they got it right, you can give some kind of reinforcement (e.g. congratulate them and tell them why they got it right). You can also plan to make some questions more difficult than others. In the first question in a series, students must find it very easy to differentiate between right and wrong answers because they need to get the basic idea before they go further. Later on in the series, you can make it more difficult to differentiate between right and wrong answers, or between examples and non-examples.
General applications:
These frame activities are for general use:
These frames were created specifically for language teaching. Some exercises are very simple but require students to respond quite quickly. This trains students to use language sponteneously as conditioned responses. Consequently, they can function at normal conversation speeds because they don't have to think long about their responses or translate between languages.
These frame types were created mainly for drills to teach foreign languages:
Frames work best when there is more than one frame on particular idea, and they progress from an easy introduction to a test on whether students have achieved the learning objective. They form a series. Here's a good, basic series for a new idea:
The final stage has a little paradox, because it has two “test” frames. They are exactly the same in that they both address the same learning objective, but they have different purposes:
This paradox also has an ethical aspect. It is normally unethical to put a task in an test for a grade if the students has never done it before. They should have had a chance to have done it and to correct misunderstanding before a final summative test.
A remedial loop is a series of frames that helps students with something they missed. You normally don't need to write any remedial loops in the first draft of your material. You'd only need to add them later on if there were things that students consistently got wrong. In either case, remedial loops have a diagnostic role. If you know when students tend to need them, you can suspect that something might not be working well in the pre-requisite course or in your course.
Pre-requisite knowledge. The first main use of remedial loops is to catch students up on pre-requisite knowledge. For example, they might have achieved a good enough grade in previous studies to get into your course, but they didn't get a perfect score. They missed something that they need for your course. A test on pre-requisite knowledge at the beginning of your course can reveal those gaps; the software can then redirect those students to a series of frames that helps them catch up on each item that they missed.
Objectives not achieved. The second main use of a remedial loop is to provide a remedy for any objectives that the student did not achieve in the formative test. This gives them the opportunity to go over the material again and achieve the objective.
In software, each incorrect answer can lead to a different remedial loop if you wish, depending on the kind of incorrect answer they choose. This kind of programming can also enable students to jump ahead if they select the correct answer.
Bloom’s taxonomy
is an excellent way to write series of frames in sequence because it relates to conceptual understanding and is good for teaching ideas, theories, approaches, and methodologies. It takes some practice to get right, but it’s brilliant when you do.
The basic sequence for teaching one rather complex idea or way of doing something, given below, is divided into the levels up to application, and the levels for higher order thinking skills.
You should plan to write a frame for each level and sublevel all marked by an asterisk below:
priming.
What?questions, without interpreting or jumping to conclusions.
What would happen if . . . ?
What else could happen?
What kind of results would you expect?
Who else would be affected by … ?
What kind of responses might you get?
If you teach in higher education, you can add a whole new level of sophistication by using these higher level thinking skills. In some kinds of topics, it's enough to teach people a set of steps on how to do X. But in higher education, you often face different approaches to a topic. For example, you might face different theoretical models of X, or a toolkit of different sets of steps on how to do Y.
However, it is much more difficult to work with predetermined responses, because students generally need to be able to create their own unique responses.
To implement these levels, you will need to teach two (or more) approaches.
lower levelsequence above from Bloom's taxonomy.
You are then ready to compare the different approaches and answer some higher-level, more theoretical questions:
These questions all require that students each give their own unique, original answers. Consequently, they cannot be automatically reinforced by comparing them to predetermined answers, so they don't work as pure programmed instruction. They work better as tasks for students to present as assignments or as topics for group discussion. But that's good too.
Cognitive load theory
Some frames are very bad and frustrate anybody who uses them. Here's a top ten of the bad list.
Only copying words. Unless the lesson objective is that students must be able to copy words, there's lots of writing and not much thinking. It's busywork and a waste of time.
Unclear examples. These confuse students and lead to wrong answers. Some people call these examples that don't examp
.
Atomistic objectives. These are objectives that specify and predetermine every tiny detail of knowledge. They are vulnerable to many problems:
railroadedor bullied into accepting your views as the only right view or the only way to do something.
Recall only. Frames ask only that students recall information. Writers often think that they are asking students to understand and apply something, but the frame really only asks students to recall something.
Bad writing. This is frustrating and confusing. The challenge is to present information simply in a form that students get the point the first time.
Nice to know. Some rules are cluttered with interesting but irrelevant things that are nice to know,
but do not help students to achieve the learning objective.
Be aware of ... Some writers say that students should be aware of
factors that affect the task. The problem is that it doesn't say what students must do. In most cases, it means that the student must be able to evaluate whther those factors are relevant and adjust their approach if they are. In a few cases, students need to be able to identify any relevant factors from the context that affect implementation.
Yes or no answers. These are not always bad, but they have limited usefulness. Use them with care.
Testing, not teaching. Some frames only test what students already know, and do not teach any new knowledge. They lack enough information for the student to work out answers.
Some frames are very impersonal. That's the value of a personal mentor by video as the information source.
What do the students need to be able to do at the end of the course? Write a profile of what you want graduates to be able to do. It will need to include supporting skills and specific knowledge. Keep goal statements brief and to the point. If a goal is longer than ten words, try to re-word it. Then check that these goals can be effectively mastered in an on-line environment.
This rationale is:
Look at the students:
Do you have any time constraints .e.g semester hours? (It is also nice to be able to tell students at the beginning of a lesson how long it will probably take.)
Why should students study your course? What pressing problems will keep them motivated? What urgent needs are they seeking to meet by doing your course?
Define the problem faced that this learning is meant to resolve. Why is it urgent? As students and virtual scenario members go through the lesson, you can progressively add problems that complicate what they are doing, and keep the tension.
Other possibilities:
b. What will be the admission requirements and pre-requisite knowledge? c. Write the course objectives d. “Rules” e. Sequence of overall course f. Time allowed (This can be important for credit courses where you have to show semester hour equivalence or tie into a class schedule) g. What will be your percentange of tolerable error for student responses. (If you assume that all students must get every answer correct, your lessons will be so simple that they will be very boring. ) Good practice is currently 80%.
The goal is to check that students can use the materials effectively, and to do so, you need to fix anything difficult in your draft.
Find a point on a continuum between the AGILE and ADDIE approaches. It's a point between between trying to get your materials perfect before field-testing and rushing to field-test with sloppy materials.
The ADDIE approach is Analyze, Design, Develop, Implement, and Evaluate. It makes several assumptions:
The AGILE approach also makes several assumptions:
Plan a cut-off point. You don't have to ensure that every student will get every response for every frame correct, because it isn't necessarily a good idea to make your materials completely free from all possible misunderstanding. (That would probably make them so boring and moron-proof they they would fail for other reasons.)
Instead the challenge is to set a cut-off point. Is it okay if 80% of students respond correctly to a question in a frame? Or is 70% or 90% better? Current good practice is 80%, but you might need to adjust it up or down in your circumstances.
First, proof-read your own work carefully.
Then, test your first draft with colleagues to eliminate the most obvious glitches.
After that, do your testing with persons of the target population who have not yet seen the materials nor had the training in any other form. People who already know the subject matter or saw previous drafts of materials won't be tripped up as if they were learning from scratch. If possible, observe them actually using the program.
If something is unclear, some will tend to stop and re-read it, which is quite observable. In some cases, they will be clearly frustrated. Others will just try to skip that bit.
Look for non-verbal clues—anything that makes them hesitate because they are confused. These will probably be:
Obviously, you will also ask for verbal feedback, but the non-verbal feedback can be more useful. Then make corrections.
While you're at it, keep an eye on the time to see how long it takes.
If a frame consistently trips students up, you face two options:
Then test your materials again on a new group of people from the target population who have never seen them before, and collate suggestions for improvement. You need to use different people each time, because they need to be working the program for the first time; they can no longer give their first impression of the meaning of the questions. If necessary, you can repeat this kind of test more widely with other new groups of people from the target population.
Decide when to stop, that is, when your materials are good enough, that is:
Well-written courses require students to concentrate intensely. Students might be able to use the materials satisfactorily, but still find the course unwelcomely difficult. This might tell you whether your cut-off percentage is too high or too low.
Decide when to stop: you can keep improving forever. Any other feedback after that can go into the next edition.
Content 18
Content 19
Content 20
Content 21
Content 22
Content 23
Content 24
Content 25
Content 26
Content 27
Content 28
Content 29