Let us count the ways. The programs were designed for several different instructional environments. First, the authors considered instructors who wanted a turnkey system — that is, one that provides a series of so-called "canned" experiments. In this case, the experiments are faithful representations of actual experiments.

The authors also accommodated students and instructors who needed to vary the basic paradigms in numerous ways, simply and conveniently. They selected an index card metaphor to enable users to understand the sequencing of parts of each experiment, and provided means by which users could made hundreds of potential variations in each experiment by altering values of numerous independent variables. Custom instructions and stimuli can be easily added, as well. All this is accomplished without the user writing a single line of code.

No. You may quit an experiment at any time, saving any data collected up to that point. Alternatively, you might use the controls on the Part cards to reduce the presentation time, number of trials, and so forth to their minimum values and turn off Part cards that involve interpolated tasks whose sole purpose is to fill a time interval with an activity.

The authors surveyed the current cognition textbooks to create a list of potential paradigms for inclusion. They then assessed the currency of the topic, the percentage of textbooks referring to the experiment, the robustness and magnitude of the reported effects, and whether an appropriate amount of data could be collected in a fairly noisy (either visually or acoustically, or both) environment within a 50-minute class period. Whenever feasible, they also assessed student interest.

The Instructor's Manual provides a variety of suggestions and recommendations for use of the software in variety of courses. You can download the Instructor's Manual from the Psychology Software Inc. web site.

The institution acquires the software licenses for the software that produces the experiments. The easiest way to do this is to download the software from the Psychology Software Inc. web site. Thirty days after first using the software, it deactivates. Once a purchase order is received, Psychology Software Inc. sends a designated individual the special code needed to reactivate the programs. The software is also available on CD-ROM.

To complement the experiments, an "intelligent" spreadsheet and graphing program, based on Excel, is included with the institutional software. Because this custom software "knows" where every independent and dependent variable is stored, students simply select from a list of X-axis and Y-axis variables they wish to plot. The software aggregates the appropriate values and generates a professional-quality graph.

The materials also include a 220-page Student's Manual, which instructors order through their bookstores. The Student’s Manual provides a context, highlights key concepts, gives a clear purpose for the experiment plus a description of the design of a default experiment (generally, a replication of a published report). It also details how to customize the paradigm by including new stimuli, instructions, or changes to settings of various parameters. Included with this manual is a copy of Manuscript Mentor, which provides suggestions for data analysis and for interpreting the outcomes of each experiment. It also gives explicit pointers to the appropriate literature. In addition, this disk provides a WYSIWYG word processor that formats an APA-style laboratory report and nearly automates adherence to the rules for citations in their reference section. Students can use the word processor on their disk for personal correspondence or for work in other courses, as well.

The programs have been run successfully on 386, 486, and Pentium-class computers running Windows 3.1x, Windows95, and Windows98. As little as 16mb of RAM is needed on older machines running Windows 3.1x.

Yes. The authors have used it successfully on a Novell network for more than 5 years. Of course, it also runs well on standalone machines.

Students need only modest experience in maneuvering a mouse through an object-oriented environment. Fortunately, most undergraduates will come to you with this knowledge, but you should be alert to apprehensions among those who need special assistance. You might suggest that novices try the Quick Start tutorial from the Marble Screen to become acquainted with the operation of the various buttons, scroll bars, and so forth. The tutorial assumes only that they can use a mouse.

We schedule simple experiments such as Well-Defined Problems as the first lab, primarily to enable students to become familiar with the operation of their computer. It may be useful to advise about proper care of the Manuscript Mentor disk; students who slide open the spring-loaded door to examine the contents are asking for trouble!

Because not all students are proficient touch-typists, wherever recall data are collected, we provide one of two ways for students to enter their reports: Where recall is used to determine the participant’s mastery of the materials and readiness for the major experiment (e.g., in Mental Models), users need to type only the first four characters of the word; the program completes the word for them. When the recall data are the object of study (as in False Memories), the program will processes each entry with a spelling checker to minimize correct typographical errors.

Some contain many cards, each a duplicate of the other, to enable students to set up multipart within-subjects experimental designs. In contrast, where the situation calls for students to study some materials, wait for a specified duration, and then be tested, there may be only three Part Cards (one each for study, interpolated activity, and testing).

They should refer to the literature that we direct them to read and consider ways that they can extend those studies using this software. After collecting data in the default experiments, students should be encouraged to explore the experimental variations built into the Part Cards and the options on the Custom Setup Card to configure entirely new experiments. Consider giving students access to these programs in the lab for their senior thesis or other independent research projects.

If the researcher uses the same filename each time the experiment concludes, the data will automatically be combined on the drive specified (network or local). If data for a particular experiment exist in several files or on multiple diskettes, from the Spreadsheet, click the File|Combine Data options on the Marble Screen to merge the data into a single file. Users cannot combine data from unrelated experiments (e.g., Mental Rotations and Attention) into a single file.

Some do. If these are needed elsewhere, change the default experiment so that the minimum number of responses are collected, and use those opportunities as practice. Alternately, use the default number of trials, and exit the program without saving data. Then restart the experiment.

Users can quit at any time and either save the data collected to that point or not. If the experiment is nearly complete, we advise saving the data, but making a note to mark the data file as being an incomplete.

In some experiments, e.g., Mental Rotations, students perform in several, independent parts. In such cases, a student might quit after performing two of the three parts, save the data, then return at another time to complete the remaining component. In other multi-part experiments, independence is not a safe assumption and, therefore, it is not possible to resume on a later day at the point where a user previously quit.

When this might be an important issue (Metamemory, Implicit/Explicit Memory, False Memories, Perceptual Aftereffects and Mental Models), users can turn off all of the Part Cards except those for the test needed. When the data are saved, the file will not contain information about the duration of the retention interval. It will indicate when the final test occurred, but the researcher must calculate the interval manually.

An easy way is to use M or F as part of the Subject ID that is requested before the experiment begins. The Subject ID is saved to the data file.

More generally, users can add any number of categorical or numeric variables to a data set in the C&P3 spreadsheet. Students can use these variables to combine data within categories, e.g., to combine all of the scores of the left-handers separately from the right-handers so that the user can plot handedness as a parameter on a graph.

Once the data are saved, the programs normally exit to the Marble Setup Screen. From there, students can access the C&P3 spreadsheet to examine, graph, and print their data.

It contains only the functions most likely to be needed by students in the target courses. Thus, it cannot compute a depreciation schedule, but it can calculate such statistics as means, medians, variances, standard deviations, and regression lines. Its beauty lies in its simplicity for student use.

Because this is only a partial implementation of Excel, it does not accept as input files that were saved in an Excel format by other spreadsheets, such as Quattro. Of course, the saved data can also be used directly in any commercial spreadsheet.

It provides for line and XY graphs, as well as 2- and 3-dimensional bar graphs. Most of the graphing options available in commercial spreadsheets are supported. The software optimizes ease of use. For example, to create a basic graph, the user must merely (1) click on the levels of independent variable displayed in one box, (2) click on the dependent variables shown in a second box, (3) then specify in the same way, in a third box, the other parameters over which the data aggregation is to occur (e.g, Women/Men, Subjects 1/2/3/4, Positive Responses/Negative Responses).

Generally, yes. Whenever possible, the software responds to mouse clicks. In many experiments, the researcher can also specify keyboard responses for users. For example, in any experiment requiring timed Yes/No or Same/Different responses), the researcher can specify any set of keyboard keys that will be appropriately recognized.

Independent tests of the high resolution timer indicate that accuracy is 10 msec or better. The experiments included here typically result in differences that are 5-10 times the resolution of this timer.