Psychology 403. Laboratory in Cognitive PsychologyFall, 2003
Experiment 1.
Discrimination of Line Length Differences
Overview
In this experiment subjects must judge which of two lines is longer, signaling their response by pressing one of two keys to indicate whether the top or bottom line looked longer. From trial to trial one of two different differences in line length will occur, one smaller and more difficult than the other. In two separate blocks of trials the instructions will vary. In one block subjects will be told to respond rapidly, while in the other block they will be told to respond accurately. The data will be examined to understand how the combination of speed and accuracy must be used together to assess the relative difficulty of different tasks.
This experiment is an introduction to the use of reaction time procedures to study cognitive processes. Under a variety of conditions, reaction time procedures are used to assess the relative difficulty of performance across different experimental conditions. Commonly, the participant is required to make a decision and indicate it by a key press (or in some cases a spoken response detected by an electronic circuit called a “voice key” which detects the signal coming from a microphone.) Two dependent measures are recorded, the reaction time, the time from the presentation of the stimulus material to the response, and the accuracy of the response. It is usually assumed that greater task difficulty will be reflected in a slower reaction time and/or lower accuracy.
We will examine the use of a reaction time procedure to assess task difficulty under conditions where the relative difficulty of two different stimulus conditions is quite clear. Participants will have to decide which of two lines is longer under two different experimental conditions. On some trials the difference in the lengths of the two lines will be smaller than on other trials. The more similar the line lengths, the more difficult the discrimination of which line is longer.
In comparing two conditions, we usually wish to determine which is more difficult. However, performance is measured by two dependent measures, reaction time (or its inverse, speed) and accuracy. Faster reaction times or greater speed indicate that a task is easier. However, the measures of reaction time and accuracy are not independent of each other. Participants often may improve accuracy by slowing down, or respond faster by becoming less accurate. For this reason, just observing reaction time or just observing accuracy across two or more levels of an independent variable is not necessarily an accurate assessment of the relative difficulty of the different levels of the independent variable. If a participant is both faster and more accurate in one condition than another, it is easy to conclude that condition is easier than the other. But, if a participant is faster, but less accurate in one condition compared to another, then the difference in difficulty between the two conditions is ambiguous. The faster reaction time suggests that the task is easier, but the lower accuracy suggests otherwise. It is possible that there is no difference, and that the participant merely traded speed for accuracy in the two different conditions. Without additional, parametric experiments, it is difficult to know how large a change in accuracy will result from a specific change in speed. The trade-off which occurs between speed and accuracy when the task difficulty remains fixed is called the speed-accuracy trade-off function.
In this experiment we will illustrate the use of reaction time procedures by examining performance under different conditions. We will manipulate two different independent variables. The first will be the difficulty of the task. Participants will be asked to determine which of two different lines is longer under two different stimulus conditions that will differ in the size of the difference in length between the two lines. We will assume (quite reasonably) that the task will be easier when the difference in the lengths of the two lines is greater. We will observe how this difficulty is reflected in each of the two different dependent measures; reaction time and accuracy. In addition, we will manipulate a second independent variable, the instructions that either urge that participants perform as rapidly as possible, or as accurately as possible. This instruction will differ for two blocks of trials. (In each block the same two line length differences will be used.) Thus we will be able to observe within each block how speed and accuracy reflect the difficulty of judging the two different pairs of lines. Comparing between blocks we will be able to see how speed and accuracy vary when the same stimuli are judged under different instructions. Since the instructions will not change the difficulty of the discrimination, performance under these two conditions for a given stimulus pair will reflect the same difficulty. So we will be able to observe how speed can be traded for accuracy, and vice versa.
Method
A trial will be initiated by the participant by pressing the space bar on the computer. A plus sign will appear in the center of the screen, as a fixation target, for 500 msec. It will then disappear and the test stimuli, a pair of lines of different length, will appear. The participant will indicate whether the longer line was on the top (by pressing the "2" key" on the numberpad) or bottom (by pressing the "8" key). The difference in the lengths of the lines will be longer in the easier condition, and shorter in the harder condition. [The actual length of the lines on the screen will depend upon the specific computer and monitor on which the experiment is run, although the relative lengths of the lines will be the same on all computers.]
The two lines will be separated vertically from each other by a fixed distance, and horizontally by one of two fixed distances. The vertical location of the pair of lines on the screen will be fixed, and will be equidistant above and below the location of the fixation point. Finally, the pair of lines will appear at one of 4 locations horizontally across the screen, spaced equally across the screen. The order of trials within a block will be randomized for each participant, and the order of blocks with different instructions will be randomized across participants.
There will be three blocks of trials, a practice block, a block in which the instructions are to respond as accurately as possible, and a block in which the instructions are to respond as quickly as possible. The practice block will always occur first, and participants will be required to reach a criterion level of performance before proceeding to the experimental conditions. The first practice block will be 14 trials, and subsequent blocks of 14 trials will be required until the criterion level of performance (80% correct) has been reached, attempting to insure that the participants understand the task and have mastered the required responses. The order of the experimental blocks will vary randomly across subjects, with some receiving the instructions to be accurate first and others receiving the instructions to respond rapidly first.
Each experimental block will consist of 96 trials, half with the easier discrimination and half with the more difficult discrimination. In addition, the longer stimulus will appear on top half of the time, and the shorter stimulus will be on top on the remainder of the trials.
Detailed instructions on how to run the experiment on the computer will be provided in class.
LineLengthF03.doc