Does transmission modality influence algorithm transmission rates?
| Autor: | Thompson, Bill, van Opheusden, Bas, Griffiths, Tom |
|---|---|
| Rok vydání: | 2022 |
| Předmět: | |
| DOI: | 10.17605/osf.io/4pr32 |
| Popis: | An experiment assessing whether people are able to learn more effectively from algorithm descriptions or algorithm demonstrations (or whether both are necessary for effective transmission). Participants will complete the the sorting task described in https://osf.io/nqkjm. Participants will be randomly assigned to one of three treatments: Description; Demonstration; Description + Demonstration. All participants will be able to select models and will be able to observe task performance of models. A power analysis along with practical considerations leads us to a target sample size of 150 participants per treatment (= 450 participants total). We will conduct separate regression analyses to assess differences between treatments in task performance and algorithm transmission rates (Gnome sort and Selection sort). |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|