How Monitor Characteristics Affect Human Perception in Visual Computer Experiments: CRT vs. LCD Monitors in Millisecond Precise Timing Research

Michaela Rohr,Alexander Wagner

doi:10.1038/s41598-020-63853-4

Abstract

Liquid crystal display (LCD) monitors are nowadays standard in computerized visual presentation. However, when millisecond precise presentation is concerned, they have often yielded imprecise and unreliable presentation times, with substantial variation across specific models, making it difficult to know whether they can be used for precise vision experiments or not. The present paper intends to act as hands-on guide to set up an experiment requiring millisecond precise visual presentation with LCD monitors. It summarizes important characteristics relating to precise visual stimulus presentation, enabling researchers to transfer parameters reported for cathode ray tube (CRT) monitors to LCD monitors. More importantly, we provide empirical evidence from a preregistered study showing the suitability of LCD monitors for millisecond precise timing research. Using sequential testing, we conducted a masked number priming experiment using CRT and LCD monitors. Both monitor types yielded comparable results as indicated by Bayes factor favoring the null hypothesis of no difference between display types. More specifically, we found masked number priming under conditions of zero awareness with both types of monitor. Thus, the present study highlights the importance of hardware settings for empirical psychological research; inadequate settings might lead to more “noise” in results thereby concealing potentially existing effects.

Highlights

The present paper summarizes the current knowledge base regarding important differences between cathode ray tube (CRT) and Liquid crystal display (LCD) monitors; it aims to provide a hands-on guide for the setup of computer experiments using Liquid crystal display monitors (LCDs) monitors in a manner that yields reliable presentation times and Cathode ray tubes (CRTs)-comparable results
Data collection was stopped after the Bayes factors (BF) reached either (a) BF01 > 6 in favor of the null hypothesis of no difference in priming effects for CRT and LCD monitors, or (b) BF10 > 6 in favor of the alternative hypothesis that there is a difference between CRT and www.nature.com/scientificreports
We present the Bayes factors based on sequential hypothesis testing as preregistered, alongside results of conventional null-hypothesis significance testing (NHST) on the final sample to allow comparison with the original results of Kunde et www.nature.com/scientificreports

Summary

Introduction

The present paper summarizes the current knowledge base regarding important differences between CRT and LCD monitors; it aims to provide a hands-on guide for the setup of computer experiments using LCD monitors in a manner that yields reliable presentation times and CRT-comparable results. A white pixel is achieved by aligning the crystals such that maximum light is allowed to pass through, until a different, non-white color needs to be displayed (see the lower panel of Fig. 1 for an LCD pixel’s brightness over time). This is a static process, not a pulsed one as in CRTs. In theory, the difference in presentation methods, namely a strobing versus a static image, should be of no consequence if the light energy that falls onto the retina remains the same over the time period of one single frame. In principle, LCD and CRT monitors should be able to yield comparable results

Objectives

Methods

Results

Conclusion