From conflicts and confusion to doubts: Examining review inconsistency for fake review detection

2021 | Decision Support Systems | Citations: 12

Authors: Shan, Guohou; Zhou, Lina; Zhang, Dongsong

Abstract: Inconsistency in online consumer reviews (OCRs) may cause uncertainty and confus ... Expand

Abstract: Inconsistency in online consumer reviews (OCRs) may cause uncertainty and confusion to consumers when they make purchase decisions. However, there is a lack of a systematic and empirical investigation of review inconsistency in the literature. This research characterizes review inconsistency from multiple aspects, including rating-sentiment, content, and language, and proposes hypotheses about their effects on fake OCR detection by drawing upon deception and attitude-behavior consistency theories. We characterize review inconsistency with 22 features, and test the hypotheses with machine learning models developed for fake OCR detection. Our empirical evaluation results using real OCRs not only confirm the presence of review inconsistency, but also demonstrate significant positive effects of review inconsistency on the performance of fake OCR detection. The research findings have important implications for improving the effectiveness of consumer decision making and the trustworthiness of OCRs. Collapse

Topics: artificial intelligence decision making systems design

Methods: sensitivity analysis machine learning descriptive statistic sentiment analysis paired t-test

Transfer Learning in Dynamic Data Environments: Trade-offs in Response to Changes

2020 | International Conference on Information Systems | Citations: 0

Authors: Peng, Jiaxu; Hahn, Jungpil

Abstract: In dynamic data environments, we often lack sufficient information to adjust pre ... Expand

Abstract: In dynamic data environments, we often lack sufficient information to adjust prediction models in a timely manner. In this study, we investigate whether and how we can use transfer learning (i.e., training a model using large but potentially less relevant data sets consisting of both historical and recent source data) when there is only a small amount of source data that exhibits the target data pattern. We provide theoretical insights on when and to what extent transfer learning works by using a sample selection perspective to represent changes. We conduct simulation analyses to examine two practical tradeoffs data analysts face with data changes – 1) whether or not they should use transfer learning to retrain the prediction model, and 2) whether they should adjust the prediction model immediately or later when more timely-relevant source data becomes available. Our study offers theoretical and practical guidelines for data analytics in dynamic data environments. Collapse

Topics: analytical information system

Methods: simulation conceptual analysis Monte Carlo simulation computational algorithm theory development

Polynomial Regression and Measurement Error: Implications for Information Systems Research

2020 | ACM SIGMIS Database | Citations: 0

Authors: Aguirre-Urreta, Miguel I.; Rönkkö, Mikko; Hu, Jiang

Abstract: Many of the phenomena of interest in information systems (IS) research are nonli ... Expand

Abstract: Many of the phenomena of interest in information systems (IS) research are nonlinear, and it has consequently been recognized that by applying linear statistical models (e.g., linear regression), we may ignore important aspects of these phenomena. To address this issue, IS researchers are increasingly applying nonlinear models to their datasets. One popular analytical technique for the modeling and analysis of nonlinear relationships is polynomial regression, which in its simplest form fits a "U-shaped" curve to the data. However, the use of polynomial regression can be problematic when the independent variables are contaminated with measurement error, and the implications of error can be more severe than in linear models. In this research, we discuss a number of techniques that can be used for modeling polynomial relationships while simultaneously taking measurement error into account and examine their performance by using a simulation study. In addition, we discuss the use of marginal and response surface plots as interpretational aides when evaluating the results of polynomial models and showcase their use through a practical example using a well-known dataset. Our results clearly indicate that the use of a linear regression analysis for this kind of model is problematic, and we provide a set of recommendations for future IS research practice. Collapse

Topics: IT skill mobile phone

Methods: simulation consistent partial least squares path modeling statistical power analysis regression analysis method polynomial regression

Detecting Common Method Bias: Performance of the Harman's Single-Factor Test

2019 | ACM SIGMIS Database | Citations: 0

Authors: Aguirre-Urreta, Miguel I.; Hu, Jiang

Abstract: Lack of careful consideration of common method effects in empirical research can ... Expand

Abstract: Lack of careful consideration of common method effects in empirical research can lead to several negative consequences for the interpretation of research outcomes, such as biased estimates of the validity and reliability of the measures employed as well as bias in the estimates of the relationships between constructs of interest, which in turn can affect hypothesis testing. Taken together, these make it very difficult to make any interpretations of the results when those are affected by substantive common method effects. In the literature, there are several preventive, detective, and corrective techniques that can be employed to assuage concerns about the possibility of common method effects underlying observed results. Among these, the most popular has been Harman's Single-Factor Test. Though researchers have argued against its effectiveness in the past, the technique has continued to be very popular in the discipline. Moreover, there is a dearth of empirical evidence on the actual effectiveness of the technique, which we sought to remedy with this research. Our results, based on extensive Monte Carlo simulations, indicate that the approach shows limited effectiveness in detecting the presence of common method effects and may thus be providing a false sense of security to researchers. We therefore argue against the use of the technique moving forward and provide evidence to support our position. Collapse

Topics: logistics management IS discipline anonymity

Methods: simulation data transformation principal component analysis factor analysis sensitivity analysis

From composites to factors: Bridging the gap between PLS and covariance‐based structural equation modelling

2019 | Information Systems Journal | Citations: 0

Authors: Kock, Ned

Abstract: Partial least squares (PLS) methods possess desirable characteristics that have ... Expand

Abstract: Partial least squares (PLS) methods possess desirable characteristics that have led to their extensive use in the field of information systems, as well as many other fields, for path analyses with latent variables. Such variables are typically conceptualized as factors in structural equation modelling (SEM). In spite of their desirable characteristics, PLS methods suffer from a fundamental problem: Unlike covariance‐based SEM, they do not deal with factors, but with composites, and as such do not fully account for measurement error. This leads to biased parameters, even as sample sizes grow to infinity. Anchored on a new conceptual foundation, we discuss a method that builds on the consistent PLS technique and that estimates factors, fully accounting for measurement error. We provide evidence that this new method shares the property of statistical consistency with covariance‐based SEM but, like classic PLS methods, has greater statistical power. Moreover, our method provides correlation‐preserving estimates of the factors, which can be used in a variety of other tests. For readers interested in trying it, the new method is implemented in the software WarpPLS. Our detailed discussion should facilitate the implementation of the method in any numeric computing environment, including open source environments such as R and GNU Octave. Collapse

Topics: open source accounting business process redesign packaged software

Methods: partial least squares regression structural equation modeling ordinary least square statistical power analysis consistent partial least squares path modeling

Theories: information systems theory

Drawing-Based Automatic Dementia Screening Using Gaussian Process Markov Chains

2018 | HICSS | Citations: 0

Authors: Lam, Max WY; Liu, Xunying; Meng, Helen ML; Tsoi, Kelvin KF

Abstract: Screening tests play an important role for early detection of dementia. Among th ... Expand

Abstract: Screening tests play an important role for early detection of dementia. Among those widely used screening tests, drawing tests have gained much attention in clinical psychology. Traditional evaluation of drawing tests totally relies on the appearance of drawn picture, but does not consider any time-dependent behaviour. We demonstrated that the processing speed and direction can reflect the decline of cognitive function, and thus may be useful for disease screening. We proposed a model of Gaussian process Markov chains (GPMC) to study the complex associations within the drawing data. Specifically, we modeled the process of drawing in a state-space form, where a drawing state is composed of drawing direction and velocity with consideration of the processing time. For temporal modeling, our scope focused more on discrete-time Markov chains on continuous state space. Because of the short processing time of picture drawing, we applied higher-order of Markov chains to model long-term temporal correlation across drawing states. Gaussian process regression was used for universal function approximation to flexibly infer the state transition function. With Gaussian process prior to the distribution of function space, we could encode high-level function properties such as noisiness, smoothness and periodicity. We also derived an efficient training mechanism for complex Gaussian process regression on bivariate Markov chains. With GPMC, we present an optimal decision rule based on Bayesian decision theory. We applied our proposed method to a drawing test for dementia screening, i.e. interlocking pentagon-drawing test. We tested our models with 256 subjects who are aged from 65 to 95. Finally, comparing to the traditional methods, our models showed remarkable improvement in drawing test for dementia screening. Collapse

Topics: tablet cloud computing digital platform Android Python

Methods: Markov chain Markov chain Monte Carlo machine learning conceptual modelling design science

Theories: decision theory

A Multicollinearity and Measurement Error Statistical Blind Spot: Correcting for Excessive False Positives in Regression and Pls

2017 | Management Information Systems Quarterly | Citations: 5

Authors: Goodhue, Dale L.; Lewis, William; Thompson, Ron

Abstract: Multiple regression has a previously unrecognized “statistical blind spot” becau ... Expand

Abstract: Multiple regression has a previously unrecognized “statistical blind spot” because when multicollnearity and measurement error are present, both path estimates and variance inflation factors are biased. This can result in overestimated t-statistics, and excessive false positives. PLS has the same weakness, but CB-SEM’s estimation process accounts for measurement error, avoiding the problem. Bringing together partial insights from a range of disciplines to provide a more comprehensive treatment of the problem, we derive equations showing false positives will increase with greater multicollinearity, lower reliability, greater effect size in the dominant correlated construct, and, surprisingly, with higher sample size. Using Monte Carlo simulations, we show that false positives increase as predicted. We also provide a correction for the problem. A literature search found that of IS research papers using regression or PLS for path analysis, 33% were operating in this danger zone. Our findings are important not only for IS, but for all fields using regression or PLS in path analysis. Collapse

Topics: anonymity website IT work spreadsheet marketing management

Methods: partial least squares regression structural equation modeling descriptive statistic simulation Monte Carlo simulation

The dynamic predictive power of company comparative networks for stock sector performance

2016 | Information & Management | Citations: 13

Authors: Chen, Kun; Luo, Peng; Xu, Dongming; Wang, Huaiqing

Abstract: As economic integration and business connections increase, companies actively in ... Expand

Abstract: As economic integration and business connections increase, companies actively interact with each other in the market in cooperative or competitive relationships. To understand the market network structure with company relationships and to investigate the impacts of market network structure on stock sector performance, we propose the construct of a company comparative network based on public media data and sector interaction metrics based on the company network. All the market network structure metrics are integrated into a vector autoregression model with stock sector return and risk. Several findings demonstrate the dynamic relationships that exist between sector interactions and sector performance. First, sector interaction metrics constructed based on company networks are significant leading indicators of sector performance. Interestingly, the interactions between sectors have greater predictive power than those within sectors. Second, compared with the company closeness network, the company comparative network, which labels the cooperative or competitive relationships between companies, is a better construct to understand and predict sector interactions and performance. Third, competitive company interactions between sectors impact sector performance in a slower manner than cooperative company interactions. The findings enrich financial studies regarding asset pricing by providing additional explanations of company/sector interactions and insights into company management using industry-level strategies. Collapse

Topics: social network big data digital news blog network effect

Methods: machine learning longitudinal research cross sectional research time series analysis support vector machine

Theories: financial theory

More Than Meets the Eye: How Oculometric Behaviors Evolve Over the Course of Automated Deception Detection Interactions

2016 | Journal of Management Information Systems | Citations: 17

Authors: Proudfoot, Jeffrey G.; Jenkins, Jeffrey L.; Burgoon, Judee K.; Jr, Jay F. Nunamaker

Abstract: Eye-tracking technology has exhibited promise for identifying deception in autom ... Expand

Abstract: Eye-tracking technology has exhibited promise for identifying deception in automated screening systems. Prior deception research using eye trackers has focused on the detection and interpretation of brief oculometric variations in response to stimuli (e.g., specific images or interview questions). However, more research is needed to understand how variations in oculometric behaviors evolve over the course of an interaction with a deception detection system. Using latent growth curve modeling, we tested hypotheses explaining how two oculometric behaviors—pupil dilation and eye-gaze fixation patterns—evolve over the course of a system interaction for three groups of participants: deceivers who see relevant stimuli (i.e., stimuli pertinent to their deception), deceivers who do not see relevant stimuli, and truth-tellers. The results indicate that the oculometric indicators of deceivers evolve differently over the course of an interaction, and that these trends are indicative of deception regardless of whether relevant stimuli are shown. Collapse

Topics: deception detection criminality

Methods: oculometry experimental group experiment latent growth curve modeling statistical hypothesis test

Theories: conflict theory

Sample Size Determination and Statistical Power Analysis in PLS Using R: An Annotated Tutorial

2015 | Communications of the Association for Information Systems | Citations: 0

Authors: Aguirre-Urreta, Miguel; Rönkkö, Mikko

Abstract: Partial least squares (PLS) is one of the most popular analytical techniques emp ... Expand

Abstract: Partial least squares (PLS) is one of the most popular analytical techniques employed in the information systems field. In recent years, researchers have begun to revisit commonly used rules-of-thumb about the minimum sample sizes required to obtain reliable estimates for the parameters of interest in structural research models. Of particular importance in this regard is the a priori assessment of statistical power, which provides valuable information to be used in the design and planning of research studies. Though the importance of conducting such analyses has been recognized for quite some time, a review of the empirical research employing PLS indicates that they are not regularly conducted or reported. One likely reason is the lack of software support for these analyses in popular PLS packages. In this tutorial, we address this issue by providing, in tutorial form, the steps and code necessary to easily conduct such analyses. We also provide guidance on the reporting results. Collapse

Topics: open source

Methods: statistical power analysis partial least squares regression simulation SEM tool descriptive statistic