Robustness of the Two-Sample T-Test

Posten, Harry O.

doi:10.1007/978-94-009-6528-7_23

Cited by 72 publications

(38 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Before carrying out the ANCOVA, we tested a number of analytic assumptions, including data normality, homogeneity of variance, linearity of the covariate and the dependent variables, and homogeneity of regression slopes (Keselman et al, 1998). Our tests of assumptions showed a violation of homogeneity of variance (F = 23.58, p < .001), but given that ANCOVA is a robust type of analysis with respect to this assumption and our large sample size (Posten, 1984;Ramsey, 1980), it was clear that the observed variance differences would not interfere with the analysis. All other tests indicated that remaining assumptions had not been violated.…”

Section: Research Question 1: How Does Peer Feedback Compare With Tramentioning

confidence: 99%

What Is the Influence of Peer Feedback and Author Response on Chinese University Students’ English Writing Performance?

Zhang

McEneaney

2019

Reading Research Quarterly

View full text Add to dashboard Cite

Enhancing writing pedagogy in university‐based courses for students of English as a foreign language has been a long‐term goal of educators and researchers. Most recent work, however, has relied on qualitative methodologies that limit our capacity to generalize findings. In this quantitative study, the authors adopted a quasi‐experimental control group design to investigate effects of peer feedback and author response on university students’ writing performance in a context of English as a foreign language. A total of 198 sophomores majoring in English took part in a 15‐week research study. The authors conducted a repeated‐measures analysis of covariance to examine the effects of a traditional teacher feedback instructional group compared with an experimental peer feedback group. Results showed that the peer feedback group showed a significantly greater writing performance gain than the traditional teacher feedback group. Furthermore, follow‐up analyses focusing on the experimental group that relied on a series of multiple regressions showed that both the quality of students’ feedback and the quality of their author responses were significant and stable predictors of students’ writing performance. Results also showed that the quality of students’ author responses had a slightly larger impact on writing performance than the quality of students’ feedback. This study provides clear evidence of the power of formative peer feedback as an approach to support learning that both empowers students and addresses widely cited problems that arise when the instructor is the only source of feedback in a writing course.

show abstract

Section: Research Question 1: How Does Peer Feedback Compare With Tramentioning

confidence: 99%

What Is the Influence of Peer Feedback and Author Response on Chinese University Students’ English Writing Performance?

Zhang

McEneaney

2019

Reading Research Quarterly

View full text Add to dashboard Cite

show abstract

“…The effect of non-normal distributions has been investigated in many studies (e.g., F statistic: Gayen 1950; t statistic: Boneau 1960, Ratcliffe 1968, Posten 1984Q statistic: Ramseyer andTcheng 1973, Brown 1974). It would appear that, if sample sizes and variances are close to equal, only severe non-normality (e.g., bimodal distributions or extreme skewness) or the presence of outliers is a problem for parametric one-way ANOV As and UMCPs.…”

Section: Assumptionsmentioning

confidence: 99%

Comparisons of Treatments After an Analysis of Variance in Ecology

Day¹,

Quinn²

1989

Ecological Monographs

1,872

1,059

View full text Add to dashboard Cite

:J.bstra.ct. ~he statistical literature on tests to compare treatments after the analysis of vanance IS reviewed, and the use of these tests in ecology is examined. Monte Carlo simulations on normal and lognormal data indicate that many of the tests commonly used are ina~propriate or inefficient. Particular tests are recommended for unplanned multiple com~ansons on the basi~ of controlling experimentwise type I error rate and providing maximum power: These mclude tests for parametric and nonparametric cases, equal and unequal sample sizes, homogeneous and heterogeneous variances, non-independent means (repeated measures or adjusted means), and comparing treatments to a control. Formulae an~ a worked example are provided. The problem of violations of assumptions, especially vanance heterogeneity, was investigated using simulations, and particular strategies are recommen?ed. The advantages and use of planned comparisons in ecology are discussed, and the philosophy of hypothesis testing with unplanned multiple comparisons is considered in relation to confidence intervals and statistical estimation.

show abstract

“…skew was <2.0 and kurtosis <9.0). 17 A P-value <0.05 (two-tailed) was considered significant. Data were analysed using SPSS version 20 (SPSS Inc., Chicago, IL, USA).…”

Section: Discussionmentioning

confidence: 99%

Take Charge of Pain: evaluating a community‐targeted self‐management education program for people with musculoskeletal pain

Hoon

Smith

Black

et al. 2016

Health Prom J of Aust

View full text Add to dashboard Cite

Issue addressed Musculoskeletal conditions are highly prevalent, affecting 28% of the Australian population. Given the persistent nature of many musculoskeletal conditions self-management is recognised as an important aspect of effective disease management. However, participant recruitment and retention for formal self-management programs is a challenge. Methods Arthritis SA (Arthritis Foundation of South Australia, a non-profit community health organisation) redesigned a shorter, community-orientated self-management education program delivered by health professionals. The program utilises aspects of the Stanford model of chronic disease self-management and motivational interviewing as well as principles of adult learning to create an effective learning environment. The program aims to guide participants to learn and practise a range of pain management strategies that are known to be effective in improving quality of life. This study used a pre- and post-test (at 6 weeks) design to determine whether this program achieved benefits in self-reported health outcomes. Outcomes that were measured included pain, fatigue, health distress, self-efficacy and communication. Results A response rate of 47% (n=102) was achieved and small but statistically significant improvements in mean [s.d.] pain scores (6.1 [2.3] to 5.4 [2.4], P=0.001), health distress (2.3 [1.3] to 2.0 [1.3], P=0.002) and self-efficacy (6.2 [2.1] to 6.8 [2.2], P=0.002) were found. Conclusion Community-based participants of this shorter, focused program recorded small but significant improvements in self-reported pain, health distress and self-efficacy. For those who completed the current program, Arthritis SA is currently exploring the potential of developing a booster session to promote sustainable positive health outcomes. So what? Supporting self-management through education is recognised as important but also as a key challenge for effective management of musculoskeletal conditions. Using a pre-post evaluation design, this study demonstrated effectiveness (short-term improvements for self-reported pain, health distress and self-efficacy) for a redesigned and shortened community-targeted program focusing on musculoskeletal pain.

show abstract

Robustness of the Two-Sample T-Test

Cited by 72 publications

References 7 publications

What Is the Influence of Peer Feedback and Author Response on Chinese University Students’ English Writing Performance?

What Is the Influence of Peer Feedback and Author Response on Chinese University Students’ English Writing Performance?

Comparisons of Treatments After an Analysis of Variance in Ecology

Take Charge of Pain: evaluating a community‐targeted self‐management education program for people with musculoskeletal pain

Contact Info

Product

Resources

About