Effect magnitude: A different focus

Kirk, Roger E.

doi:10.1016/j.jspi.2006.09.011

Cited by 83 publications

(59 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…That is, even as more people use the Internet to gather information, the credibility of that information remains uncertain and often depends on the information sources, such as personal websites or social networks (e.g., Flanagin and Metzger, 2000, 2007Johnson and Kaye, 2009;Westerman et al, 2012). To assess the credibility of a blog for example, consumers often consider factors such as the source's knowledge, transparency, and reliability or the consistency and accuracy of the content (Kang, 2010).…”

Section: Perceived Credibilitymentioning

confidence: 99%

Do online hotel rating schemes influence booking behaviors?

Casaló

Flavián

Guinalíu

et al. 2015

International Journal of Hospitality Management

245

148

View full text Add to dashboard Cite

Section: Perceived Credibilitymentioning

confidence: 99%

Do online hotel rating schemes influence booking behaviors?

Casaló

Flavián

Guinalíu

et al. 2015

International Journal of Hospitality Management

245

148

View full text Add to dashboard Cite

“…The increase in the number of criticisms has been associated with the growing recognition of the limitations associated with using significance tests as the only criterion for interpreting the significance of the results (Batanero 2000;Falk 1998;Kirk 2007;Krueger 2001;Nickerson 2000).…”

mentioning

confidence: 99%

Uses and abuses of statistical significance tests and other statistical resources: a comparative study

2010

View full text Add to dashboard Cite

The empirical study we present here deals with a pedagogical issue that has not been thoroughly explored up until now in our field. Previous empirical studies in other sectors have identified the opinions of researchers about this topic, showing that completely unacceptable interpretations have been made of significance tests and other statistical resources. We tried to show the degree to which these errors of interpretation appear in a sample of university professors and researchers. To evaluate the beliefs and opinions of the university professors about certain methodological and statistical questions, we adapted the "Psychometrics Group Instrument" questionnaire by Mittag (1999) and Mittag and Thompson (Educational Researcher 29:14-20, 2000). We specifically wanted to detect opinions about: (1) avoiding inappropriate interpretations, (2) rectifying incorrect uses, (3) creating a debate on possible alternatives, and (4) proposing editorial changes in the publication criteria. In this study, we compare the results we obtained with those obtained by two other equivalent studies (Gordon, Journal of Vocational Educational Research 26:1-18, 2001; Mittag and Thompson, Educational Researcher 29:14-20, 2000). The characteristics of the sample and the scope of the results are described, and a critical reflection is carried out on typical misunderstandings in interpretation that show a worrisome lack of understanding of some basic methodological and statistical concepts.Résumé L'étude empirique que nous présentons traite d'une question pédagogique qui n'a pas été explorée à fond jusqu'à maintenant dans notre domaine. Études empiriques Eur

show abstract

“…This result, given the fact that BMLP11-2a's median AUROC is already very high, may be practically significant, i.e. may be useful in practice [40]. Note that practical significance is different from (the often criticized notion of) statistical significance, which gives the probability of obtaining similar or more extreme data if the null hypothesis is true.…”

Section: Resamplingmentioning

confidence: 91%

Using Bayesian neural networks with ARD input selection to detect malignant ovarian masses prior to surgery

Calster

Timmerman

Nabney

et al. 2007

Neural Comput Applic

View full text Add to dashboard Cite

In this paper, we applied Bayesian multi-layer perceptrons (MLP) using the evidence procedure to predict malignancy of ovarian masses in a large (n = 1,066) multicentre data set. Automatic relevance determination (ARD) was used to select the most relevant inputs. Fivefold crossvalidation (5CV) and repeated 5CV was used to select the optimal combination of input set and number of hidden neurons. Results indicate good performance of the models with area under the receiver operating characteristic curve values of 0.93-0.94 on independent data. Comparison with a linear benchmark model and a previously developed logistic regression model shows that the present problem is very well linearly separable. A resampling analysis further shows that the number of hidden neurons specified in the ARD analyses for input selection may influence model performance. This paper shows that Bayesian MLPs, although not frequently used, are a useful tool for detecting malignant ovarian tumours.

show abstract

Effect magnitude: A different focus

Cited by 83 publications

References 78 publications

Do online hotel rating schemes influence booking behaviors?

Do online hotel rating schemes influence booking behaviors?

Uses and abuses of statistical significance tests and other statistical resources: a comparative study

Using Bayesian neural networks with ARD input selection to detect malignant ovarian masses prior to surgery

Contact Info

Product

Resources

About