Problem Solving Tasks

tasks. Its main difference to game is that it does not create trials on the fly but each single task has to exactly correspond to a single trial. The problems involved my have some internal step-like structure but the single steps are not resolved by prob. Currently there are only 2 types of problems available: anagram rearrangement and labyrinth problems.

Program Features

Anagram Problems

An anagram is a string anagram of characters which have to be rearranged in order to convert the string into a word. Rearrangement is done by sequentially pointing to a single character of the test string with the mouse cursor and press a mouse button. The character then is transfered from the test string to the next free position of the solution string answer. Pointing to a character in the solution string moves the character back to its original position in the test string. The trial is considered to be finished by prob if all characters of the test string have been moved to the solution string. The program may check for correct solutions by comparing the solution string with the string parameter word and setting eval accordingly.

Labyrinth Problems

A labyrinth is defined by a series of connected straight line segments. Each labyrinth is contained in a file whose name is defined by labfile. The labyrinth files are searched in the directory given by labdir. A single line in the file contains points of a single polygon which have to be connected. Each new line starts a new polygon. The single polygon points are given as xy-coordinate pairs. Before drawing the polygons its coordinates are transformed such that the wole labyrinth is centered on the screen. Labyrinth size is controlled by labunit which gives the number of PXL drawing units (pxlm) for each unit of the labyrinth coordinates in the file. Figure lab shows an example of a labyrinth file and its corresponding labyrinth. In addition to the polygon points the first line in the file contains the start and the end position for the problem. Each labyrinth may contain up to maxpolygons polygons and each single polygon may contain up to maxppoints points.

Insert Figure lab here.

The subject starts with the mouse pointing to the start position. Imagine an extensible rubber string connected to the mouse pointer. The string may be fixed to a given position by pressing the left mouse button. Initially the string is fixed only to the start position. The subject has to move through the labyrint and fix the string such that it never contacts any of the labyrinth walls. The trial is finished if the string is fixed near the end position.

Faked Solution Time

During each session the program computes the average solution time for the single problems. This time is stored in timeaverage. This time may be fed back to the subject by messages between trials or blocks. The feedback messages may be biased by using faketime instead of timeaverage. The actual value of faketime is timeaverage multiplied by fakefactor.

Response Evaluation and Feedback

The parameter eval contains the result of the response evaluation. According to this result the feedback message may be one of the following string parameters: corrmsg, falsemsg, timeoutmsg, or illegalmsg. The default is feedbackmsg. The feedback message is shown for feedbacktime ms.

Program Parameters

General Parameters

foreground (int)
Foreground objects drawing color.

background (int)
Background color.

msgcolor (int)
Text message color.

question (string)
Question between trials. If this string is nonzero then it will be asked as a question between trials.

qresponse (int)
Response to question.

task (int)
Type of problem to solve:

Anagram task.

Labyrinth task.

estimatemsg (string)
Message text that requests an estimate of the solution time.

fakefactor (float)
Fake feedback time factor.

faketime (float)
Fake average solution time.

flags (string)
These flags are available:

Present a request for the expected solution time from the subject before starting a trial. The requesting message is estimatemsg and the response is stored in response.

Ask for a motivation scale value before starting a trial. The question asked is question and the response is stored in qresponse. The response is collected by presenting a rating scale on the screen and allowing the subject to select a position on the scale.

If this flag is set then the rubber string may not be fixed at any arbitrary position but is restricted to positions exactly between the labyrinth walls. The point automatically snaps on to the grid of restricted points.

timeaverage (float)
Correct average solution time until now.

timeunit (int)
Timing unit in ms. Most time values for prob are not given in milliseconds but in units of timeunit milliseconds. This affects: dsptime, rtime, and all values derived from these like timeaverage and faketime.

waitclear (int)
Waiting time before clear screen.

Anagram Task Parameters

anagram (string)
Anagram string.

word (string)
Solution word.

rowdist (int)
Vertical baseline distance of anagram and solution word.

coldis (int)
Horizontal character distance.

Labyrinth Task Parameters

startcolor (int)
Starting point mark color.

trackcolor (int)
Tracking rubber band color.

finishcolor (int)
Goal point mark color.

labcolor (int)
Labyrinth wall color.

labfilepath (string)
Name of the directory where to find the labyrinth files.

labunit (int)
Labyrinth coordinates unit size in pxlm.

labfile (string)
Labyrinth file name.

maxpolygons (int)
Maximum number of polygons in any single labyrinth.

maxppoints (int)
Maximum number of points per polygon found in any labyrinth.

maxtrackpoints (int)
Maximum number tracking points allowed.

intersections (int)
Number of intersections of the tracking line with the labyrinth walls generated by a subject's path.

Example Experiments

This is an example of the anagram task.

Parameter file anagram.x from directory \pxl\app\prob

Back to table of contents

Author: Hans Irtel