NetLogo User Community Models

(back to the NetLogo User Community Models)

Download If clicking does not initiate a download, try right clicking or control clicking and choosing "Save" or "Download".(The run link is disabled for this model because it was made in a version prior to NetLogo 6.0, which NetLogo Web requires.)

WHAT IS IT?

OSOSS stand for online self-organizing social system. The phrase was coined by Wiley and Edwards in their 2002 paper regarding the way in which large groups of people manage to support each others' learning needs via the network.

HOW IT WORKS

The OSOSS model is meant to model scenarios like the following:

Learners interested in Linear Algebra go to the internet and search for "linear algebra" on Google, eventually arriving at the collection of linear algebra materials in MIT's OpenCourseWare collection. Learners browse the materials available there, but occasionally come across materials too difficult for them. At this point learners may choose to post a question on Open Learning Support (a discussion forum integrated into the OCW collection where the learners may go for help).

Learners' progress through the materials will stop until either (1) their question is answered, or (2) their patience runs out. We may imagine three types of answers being provided. First, questions in OLS may be answered legitimately by other learners with (1) sufficient understanding of the material and (2) sufficient willingness to be helpful. Next, phoney answers may be provided by "show-offs" without sufficient understanding of the material but (2) with a desire to appear to know the answer. Finally, a tutor who answers all questions which go for a certain period without an answer (e.g., 2 days) may provide an answer. [Note: Such tutors are not currently available in OLS forum described in the example.]

WHAT YOU'RE SEEING

Learners (red arrowheads) visit a variety of online learning resources (blue squares). If they're unable to learn the material on their own, they ask a question and wait for an answer (the blue square turns yellow). Either their question gets answered, in which case they move on (and the square turns blue again) or they give up and move on (in which case the square turns also blue again).

HOW TO USE IT

Quick Start: Press the Setup button. Press the Go! button. Watch the pretty graphs!

This section explains how to use the sliders and buttons in the interface tab to configure and run the OSOSS model.

Learning Objects -- Use this slider to set the number of learning objects which will be placed in the space. Learning objects are implemented as patches which can be either blue (normal) or yellow (indicating that one or more learners are waiting on the patch with a question).

Learners -- Use this slider to set the number of learners which will be placed in the space. Learners are implemented as turtles which are red.

AvgInitialStudentKnows -- Learner knowledge is represented as a number between 1 and 100. Use this slider to set the average amount known by learners participating in the OSOSS. The specific amount known by each individual in the group will be distributed normally across the entire group with this slider value as the mean and a standard deviation of 15.

AvgLearningObjectDiff -- Learning object difficulty is represented as a number between 1 and 100. Use this slider to set the average difficulty of learning objects in the learning network. The specific difficulty of each learning object in the group will be distributed normally across the entire group with this slider value as the mean and a standard deviation of 15.

AvgHelpfulness -- Helpfulness is a learner's probability of attempting to go answer a question when it sees that a question has been asked. This is implemented in two ways: first, as the probability that a learner will actively look for yellow patches as opposed to making a random walk of the learning object space, and second, as the probability that a learner with sufficient understanding will actually provide an answer when it lands on a yellow patch. Helpfulness is represented as a number between 1 and 5. The specific amount of Helpfulness of each individual in the group will be distributed normally across the entire group with this slider value as the mean and a standard deviation of 2.

AvgShowOff -- ShowOff is a learner's probability of answering a question they are not qualified to answer. Showoff is represented as a number between 1 and 5. The specific amount of ShowOff of each individual in the group will be distributed normally across the entire group with this slider value as the mean and a standard deviation of 2.

AvgPatience -- Patience is how many days a learner will wait on a yellow patch for an answer to their question. Patience is represented as a number between 1 and 5. The specific amount of Patience of each individual in the group will be distributed normally across the entire group with this slider value as the mean and a standard deviation of 2.

Tutor? (On/Off) -- This switch determines whether or not a competent Tutor who will correctly answer each and every question is availble. System behavior observed with Tutor? turned on may not scale well in the real world, due to the cost of providing the service.

DaysTutorWaits -- If Tutor? is turned on, this slider tells the Tutor how long to let questions go before answering them.

THOUGHTS ON THE VALIDITY OF THE MODEL

There are several challenges to the validity of every model. Here is my current best thinking about the challenges.

Epistemology of the model:

- Learner knowledge is totalized as a value between 1 and 100.
[Hey, you've got to model it somehow.]

- Learners "know more" (+3 to knows score) after they visit a learning object whose difficulty score is higher than but within 5 points of their knows score.
[If the content isn't too far from their prior knowledge they should be able to learn somthing useful from it.]

- Learners "know more" (+2 to knows score) after the tutor answers their question. (2 points awarded to each learner waiting on the yellow patch.)

- Learners "know more" (+5 to knows score) after another learner answers their question. (5 points awarded to each learner waiting on the yellow patch.)
[These relative scores are set with Wertch's notion of the accessibility of authoritative discourse in mind.]

- Learners "know less" (-3 to knows score) after a showoff answers their question. (3 points deducted from each learner waiting on the yellow patch.)
[This score is based on Click and Clack's notion that two stupid people actually know less than either did individually.]

Pragmatics of the model:

- Learners aren't always online, they come and go.
[I think the fact that turtles spend more time on black patches than blue or yellow ones models this satisfactorily.]

- Learners don't do random walks over resource spaces.
[I've really puzzled over this one. If we assume that each learner's initial placement near a learning resource is equivalent to Google dropping them off near something they're interested in, then we make out ok.

- Due to the randomness of the walk, learners end up exploring only nearby resources in the short term.
[Can small distances on the Netlogo canvas be thought of as short link paths on the network (learning objects on the same site or linked to from a site), so that physical distance in Netlogo models a kind of semantic distance? If so, then this part of the model does fairly well. If not...

THINGS TO TRY

1. Try to maximize the questions posed to questions answered ratio. Use the Posed / Answered graph for feedback on how you are doing. What's the best you can do WITHOUT the tutor in approximately 180 days?

2. Try to maximize the speed with which learners learn. Watch the slope of the Avg Student Knows graph for feedback on how you are doing. What's the best you can do WITHOUT the tutor in approximately 180 days?

EXTENDING THE MODEL

It would be interesting to consider extending the model with flocking behavior or more advanced behavior based on learner presence. In the current model a learner only notices the existence of nearby questions waiting to be answered - s/he has no access to trails or paths created by other learners, which might allow additional organization to emerge.

RELATED MODELS

This model is related to the Termite model (where termintes find wood chips and manipulate them) and the Slime model (where bits of slime mold follow each other around). Both are in the default Netlogo model library.

CREDITS AND REFERENCES

All portions of this model except the be-helpful procedure were developed by David Wiley (http://wiley.ed.usu.edu/) and are redistributed under the terms of the Creative Commons Attribution-ShareAlike license. See http://creativecommons.org/licenses/by-sa/2.0/

The be-helpul procedure was adapted from Uri Wilensky's Slime.nlogo and is redistributed under the terms of the copyright notice below.

; *** NetLogo Model Copyright Notice ***
;
; This model was created as part of the project: CONNECTED MATHEMATICS:
; MAKING SENSE OF COMPLEX PHENOMENA THROUGH BUILDING OBJECT-BASED PARALLEL
; MODELS (OBPML). The project gratefully acknowledges the support of the
; National Science Foundation (Applications of Advanced Technologies
; Program) -- grant numbers RED #9552950 and REC #9632612.
;
; Copyright 1998 by Uri Wilensky. All rights reserved.
;
; Permission to use, modify or redistribute this model is hereby granted,
; provided that both of the following requirements are followed:
; a) this copyright notice is included.
; b) this model will not be redistributed for profit without permission
; from Uri Wilensky.
; Contact Uri Wilensky for appropriate licenses for redistribution for
; profit.
;
; This model was converted to NetLogo as part of the project:
; PARTICIPATORY SIMULATIONS: NETWORK-BASED DESIGN FOR SYSTEMS LEARNING IN
; CLASSROOMS. The project gratefully acknowledges the support of the
; National Science Foundation (REPP program) -- grant number REC #9814682.
; Converted from StarLogoT to NetLogo, 2000. Updated 2003.