Posts about reproducible paper (old posts, page 17)

The Solution to Science's Replication Crisis

The solution to science's replication crisis is a new ecosystem in which scientists sell what they learn from their research. In each pairwise transaction, the information seller makes (loses) money if he turns out to be correct (incorrect). Responsibility for the determination of correctness is delegated, with appropriate incentives, to the information purchaser. Each transaction is brokered by a central exchange, which holds money from the anonymous information buyer and anonymous information seller in escrow, and which enforces a set of incentives facilitating the transfer of useful, bluntly honest information from the seller to the buyer. This new ecosystem, capitalist science, directly addresses socialist science's replication crisis by explicitly rewarding accuracy and penalizing inaccuracy.

Reproducibility in wireless experimentation: need, challenges, and approaches

Wireless networks are the key enabling technology of the mobile revolution. However, experimental mobile and wireless research is still hindered by the lack of a solid framework to adequately evaluate the performance of a wide variety of techniques and protocols proposed by the community. In this talk, I will motivate the need for experimental reproducibility as a necessary aspect for healthy progress as accepted by other communities. I will illustrate how other research communities went through similar processes. I will then present the unique challenges of mobile and wireless experimentation, and discuss approaches, past, current, and future to address these challenges. Finally, I will discuss how reproducibility extends to mobile and wireless security research.

AN INTERNATIONAL INTER-LABORATORY DIGITAL PCR STUDY DEMONSTRATES HIGH REPRODUCIBILITY FOR THE MEASUREMENT OF A RARE SEQUENCE VARIANT

This study tested the claim that digital PCR (dPCR) can offer highly reproducible quantitative measurements in disparate labs. Twenty-one laboratories measured four blinded samples containing different quantities of a KRAS fragment encoding G12D, an important genetic marker for guiding therapy of certain cancers. This marker is challenging to quantify reproducibly using qPCR or NGS due to the presence of competing wild type sequences and the need for calibration. Using dPCR, eighteen laboratories were able to quantify the G12D marker within 12% of each other in all samples. Three laboratories appeared to measure consistently outlying results; however, proper application of a follow-up analysis recommendation rectified their data. Our findings show that dPCR has demonstrable reproducibility across a large number of laboratories without calibration and could enable the reproducible application of molecular stratification to guide therapy, and potentially for molecular diagnostics.

Reprowd: Crowdsourced Data Processing Made Reproducible

Crowdsourcing is a multidisciplinary research area in-cluding disciplines like artificial intelligence, human-computer interaction, database, and social science. One of the main objectives of AAAI HCOMP conferences is to bring together researchers from different fields and provide them opportunities to exchange ideas and share new research results. To facilitate cooperation across disciplines,repro-ducibilityis a crucial factor, but unfortunately it has not got-ten enough attention in the HCOMP community.

Vive la Petite Différence! Exploiting Small Differences for Gender Attribution of Short Texts

This article describes a series of experiments on gender attribution of Polish texts. The research was conducted on the publicly available corpus called "He Said She Said", consisting of a large number of short texts from the Polish version of Common Crawl. As opposed to other experiments on gender attribution, this research takes on a task of classifying relatively short texts, authored by many different people. For the sake of this work, the original "He Said She Said" corpus was filtered in order to eliminate noise and apparent errors in the training data. In the next step, various machine learning algorithms were developed in order to achieve better classification accuracy. Interestingly, the results of the experiments presented in this paper are fully reproducible, as all the source codes were deposited in the open platform Gonito.net. Gonito.net allows for defining machine learning tasks to be tackled by multiple researchers and provides the researchers with easy access to each other’s results.