Signal-to-Noise Ratios, Performance Criteria, and Transformations

Box, George E. P.; Shoemaker, Anne C.; Tsui, Kwok‐Leung; León, Ramón V.; Parr, William C.; Nair, Vijayan N.; Pregibon, Daryl; Carroll, Raymond J.; Ruppert, David; Gunter, Bert; Ullman, Neil R.

doi:10.2307/1270311

Cited by 242 publications

(87 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, the table lists only specific theories; we removed high-level families of theories (e.g., "organizational theory", "economic theory", and "social theory") from the results. The rightmost column in the table shows each theory's standardized moment (or signal-to-noise ratio), which measures the extent to which each theory's usage in North American IS journals was stable over time; higher values indicate greater levels of stability (Box, 1988). We discuss the stability of the various theories in this section after discussing theory frequencies.…”

Section: Research Question 1: the Theoretical Core Of North American mentioning

confidence: 99%

“…As a means of providing a methodologically sound basis through which we could compare the relative stability of the various theories, we computed the overall standardized moment (or signal-to-noise ratio) for each theory (Box, 1988). Table 3 lists the top 10 percent of all of the theories considered in the analysis ranked according to the stability of their usage in leading North American IS journals over time as measured by each theory's standardized moment.…”

Section: Figure 3 Comparative Overall Impact Of Theories By Ordinal mentioning

confidence: 99%

See 1 more Smart Citation

Theory in North American Information Systems Research: A Culturomic Analysis

Soper¹,

Turel²

2016

CAIS

View full text Add to dashboard Cite

Abstract:Since its inception, North American information systems (IS) research has relied on a broad and varied collection of theories. The core of this theoretical landscape is an important determinant of the IS research community's identity, and, as such, researchers have discussed it extensively in recent years. Nevertheless, we know few concrete facts about the composition, consistency, or evolution of this theoretical core over time. Using a set of 318 theories in conjunction with n-gram analyses, we address these issues empirically by computationally analyzing the complete text of every research paper published in three leading North American IS journals over a 24-year period. In examining these 2,215 papers and more than 3.54 billion n-gram records, we identify the theories that constitute the overall core of North American IS research and provide insights into the evolution of that core. We further identify and quantify the nature of theoretical pluralism in North American IS research and examine the evolution of the theoretical density of IS research studies over time. Finally, our results shed light on the patterns of theory co-occurrence in North American IS research studies and demonstrate how such information can facilitate increasingly imperative efforts aimed at theory consolidation and generalization.

show abstract

Section: Research Question 1: the Theoretical Core Of North American mentioning

confidence: 99%

Section: Figure 3 Comparative Overall Impact Of Theories By Ordinal mentioning

confidence: 99%

Theory in North American Information Systems Research: A Culturomic Analysis

Soper¹,

Turel²

2016

CAIS

View full text Add to dashboard Cite

show abstract

“…However, the limitations of an analysis based on the "signal to noise" ratio have been pointed out by many researchers (Box, 1988;Barreau, Chassagnon, Kobi, & Seibilia, 1999). Various methods of analysis have been proposed in robust parameter designs.…”

Section: Introductionmentioning

confidence: 99%

Analysis of robust parameter designs

Mak¹,

Nebebe²

2017

J. Mod. App. Stat. Meth.

View full text Add to dashboard Cite

show abstract

“…The following plots of response profiles indicate almost the same results of Taguchi Methods (note: B1 and B3are about the same performance by Taguchi Methods). This proves that SM ratio of Taguchi Methods is equivalent to the analysis of data using sigma ( a ) as the response[4]. Response profiles of the simple model…”

mentioning

confidence: 60%

“…However, it ignores the interaction between factors by assuming linear model of the factors, and the predicted optimal condition usually is not reproducible if the system involves with factors that are strongly interacted with each other [3,4].…”

Section: Introdliltionmentioning

confidence: 99%

Robust design for VLSI process and device

Chen

2001 6th International Workshop on Statistical Methodology (Cat. No.01TH8550)

View full text Add to dashboard Cite

This paper presents a comprehensive study of the design-of-experiments techniques for the robust design of VLSK process and device. A device or a process is said to be robust when its performance is very insensitive to the change or variation of process conditions. The focus is on the utilization of orthogonal arrays to design the experiments, and the reduction of the variation (or standard deviation) of the responses to get the robust process condition. Many useful methodologies of design-of-experiments are presented in this paper and three examples of case study are used to illustrate the procedure of robust design for VLSI process and device.

show abstract

Signal-to-Noise Ratios, Performance Criteria, and Transformations

Cited by 242 publications

References 35 publications

Theory in North American Information Systems Research: A Culturomic Analysis

Theory in North American Information Systems Research: A Culturomic Analysis

Analysis of robust parameter designs

Robust design for VLSI process and device

Contact Info

Product

Resources

About