Cognitive Modeling Signal detection theory and Jasp

I need to find a real dataset that I can run signal detection theory model on it in jasp, and then write a report. The goal of the report is for you to find your own data set, and analyze it using a cognitive model in JASP. Your report should have four sections:

Abstract. One paragraph. A summary of the entire paper.

Introduction. One or two paragraphs. Explain the context in which the data were collected, and why they are interesting or useful. This might involve the research question if they were collected in an experiment, or a more general account of what they are used for if they are naturally-occurring data. Say why the model that will be applied is a relevant one for these data.

Data. One or two paragraphs. Describe the format of the data in a precise way.

Results. Three or four paragraphs, plus figures or tables as appropriate. Apply the model and present the results. Use subsections if appropriate. Interpret each of the results in terms of the question the analysis was intended to answer (e.g., work through the results for each of the parameters in the model systematically).

Discussion. One or two paragraphs. Relate the results in an integrated way back to what is interesting and important about the data, as presented in the Introduction.

The grading rubric is that the components are worth: Abstract (5%), Introduction (10%), Data (25%), Results (50%), and Discussion (10%), so the emphasis is on explaining the data clearly and doing good analyses with well-explained results.

Any data set that involves people’s behavior, and is appropriately analyzed by the model you choose, is acceptable. Here are some possibilities

Sometimes journal papers or book chapters provide the required data in a table in the publication itself.

Often researchers make data from their experiments online. It may be posted on their web pages, or in a repository like the Open Science Framework (OSF).

T

here

are repositories of interesting data sets, some of which involve human behavior and would be appropriate at data science sites like

Kaggle

(you may need to set up a free account to access data).

It is often possible to “scrape” or otherwise collect data directly from an online source.

It is important that the data are available in the format assumed by the high threshold, SDT, or Thurstone models. This is basically:

High-threshold and signal detection theory models: Any situation where there are multiple trials of people detecting some sort of “signal” from some sort of “noise”, and information about whether the stimulus really was signal or noise

Thurstone model: Complete rankings of the same set of items provided by different people

You could also use the online sellers or cake baking models if you wanted, although they are not really core cognitive models

You only have to analyze one data set in terms of one model. But, you could analyze multiple data sets in terms of the same model (as in the Amyloid data, for which we had separate analyses for positive and negative patients). You could also analyze the same data sets with multiple models (as in the analysis of pilot landing data by both high-threshold and signal detection theory models). But you should only attempt what is feasible in the time and space constraints. One good data set appropriately analyzed by one model is certainly enough to get a perfect grade.

Whatever data you decide to use, you should prepare a .csv file (e.g., using Google Sheets, Excel, or just a text editor) to get your data set organized. Then you can use one of the .jasp model files from the course to analyze the data. The JASP cheatsheetLinks to an external site. should be helpful.

We Are Using LaTeX

The scientific report must be prepared using the LaTeX (pronounced “LAH-tek” or “LAY-tek”) document preparation system in the web-based interface provided by Overleaf. You have free access to the professional edition of Overleaf via your UCI email. There is information here. LaTeX has been around since the 1980s, and is used to prepare most published articles and books in the scientific literature.

There is a LaTeX template with some minimal instructions regarding how to set up sections, what goes in each section, how to cite references, and how to produce figures and tables, to get you started. It is linked below. Overleaf provides good online help here. There are also many LaTeX resources on the web. Just typing questions into Google usually quickly provides links to resources to solve LaTeX problems. You can also post to the Canvas discussion board.

You will not be graded on your use of LaTeX, but this is an opportunity to get some familiarity with an important scientific document preparation system, if you are not already familiar with it.

The LaTeX template you should use to write the report is here. Under the menu button in the top left of overleaf is a word count option. The maximum number of words for the report is 2500. The goal is 2000–2500 words. We may ask for a link to your Overleaf file to verify the word count if necessary.

Overleaf lets you download a .pdf of your final document. You need to submit this .pdf in Canvas, AS WELL AS your .jasp file(s).