Toward a More Nuanced Understanding of the Statistical Properties of a Median Split

Iacobucci, Dawn; Posavac, Steven S.; Kardes, Frank R.; Schneider, Matthew J.; Popovich, Deidre

doi:10.2139/ssrn.2663427

Cited by 64 publications

(85 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The medians were obtained from participants’ average power and average asymmetry calculated across the three isokinetic test velocities. The authors recognize that the sample size is marginal for this analysis and using the median to dichotomize participants may reduce effect sizes and statistical power, and increase the risk for Type II error (Cohen 1983, Iacobucci, Posavac et al 2015), hence we consider this analysis exploratory. Because groups differed by gender and mass these variables were then tested as covariates.…”

Section: Methodsmentioning

confidence: 99%

Knee extensor power asymmetry is unrelated to functional mobility of older adults

LaRoche

Villa

Bond

et al. 2017

Experimental Gerontology

View full text Add to dashboard Cite

Purpose To determine whether knee extensor power asymmetry limits functional mobility of older adults who possess muscle weakness. Methods Knee extensor power was measured in 36 older men and women (76.0 ± 7.6 yr), for each leg, on an isokinetic dynamometer at 60, 180, and 300 deg s−1 and power asymmetry was calculated as the percent difference in power between strong and weak legs, at each isokinetic velocity. 400-m walk, stair ascent, and five-repetition chair rise tests were performed to assess functional mobility. Pearson correlations were used to examine the relationship between weak leg power, strong leg power, power asymmetry, and 400-m walk time, stair ascent time, and chair rise time. Participants were then stratified into Low Power-High Asymmetry (LP-HA), Low Power-Low Asymmetry (LP-LA), High Power-High Asymmetry (HP-HA), and High Power-Low Asymmetry (HP-LA) groups who were compared for functional mobility. Results Knee extensor power asymmetry was unrelated to 400-m walk time (r = 0.16, p = 0.180), stair ascent time (r = 0.22, p = 0.094), or chair rise time (r = 0.03, p = 0.437), whereas weak and strong leg powers were equally associated with 400-m time (r = −0.62, p < 0.001; r = −0.62, p < 0.001), stair ascent time (r = −0.55, p < 0.001; r = −0.57, p < 0.001), and chair rise time (r = −0.28, p = 0.048; r = −0.31, p = 0.032), respectively. Power asymmetry was lowest at 60 deg s−1 (12%), and increased with contraction velocity (p = 0.001) to 15% at 180 deg s−1 and to 20% at 300 deg s−1. LP-HA exhibited 26% slower 400-walk time than HP-LA (p = 0.015) and 19% slower than HP-HA (p = 0.037). LP-HA had 31% slower stair ascent time than HP-LA (p = 0.033). There were no differences in chair rise performance between groups. Conclusions Knee extensor power asymmetry was unrelated to 400-m walk, stair ascent, and chair rise performance in older adults. Weak and strong limb powers were equally related to these functional measures, but a leftward shift of the power-mobility curve exists for the weak leg that may hinder clinical assessment if strength or power is measured on a single limb and symmetry is assumed. The greatest degree of knee extensor power asymmetry occurred at the fastest isokinetic velocity, which suggests high-speed muscle contractions may better differentiate laterality of function in older individuals.

show abstract

Section: Methodsmentioning

confidence: 99%

Knee extensor power asymmetry is unrelated to functional mobility of older adults

LaRoche

Villa

Bond

et al. 2017

Experimental Gerontology

View full text Add to dashboard Cite

show abstract

“…This problem was realized by a number of researchers in the consumer psychology and the broader psychology community and a number of articles on this issue were written. For a recent dialogue on when to dichotomize variables, see Pham (2015); Iacobucci, Posovac, Kardes, Schneider, and Popovich (2015a, 2015b); Rucker, McShane, and Preacher (2015); McClelland, Lynch, Irwin, Spiller, and Fitzsimons (2015). Early discussions on the subject were provided by Jaccard, Turrisi, and Wan (1990); Aiken and West (1991); and Cohen (1983), and followed up by Irwin and McClelland (2001, 2003) and MacCallum, Zhang, Preacher, and Rucker (2002).…”

Section: Understanding Spotlight Analysis: Highlighting Common Confusionmentioning

confidence: 99%

A clearer spotlight on spotlight: Understanding, conducting and reporting

Krishna

2016

J Consum Psychol

View full text Add to dashboard Cite

There has been a remarkable increase in the use of spotlight analysis to examine any interactive effect between an independent variable and a continuous moderator. Most of the spotlight analyses have been conducted at one standard deviation above and below the mean value of the moderator, even when alternate methods are more appropriate. Additionally, many spotlight analyses are not conducted correctly. More importantly, results for spotlight analyses are reported in a manner that makes it virtually impossible for mistakes to be detected. This article focuses on "understanding," "conducting," and "reporting" spotlight analyses. By posing questions for the reader, it highlights some common mistakes made when doing spotlight analysis and explains why confusion often arises. Then it provides an easy to understand way to do spotlight analysis for some popular contexts. Alternatives to spotlight analysis are also briefly discussed. Finally, it suggests how to report results for spotlight analysis and for the alternatives. Pointing out recurrent mistakes should prevent perpetuation of misleading practices. Similarly, reporting essential details of the analyses should prevent mistakes from going undetected.

show abstract

“…To investigate individual differences in learning, we performed a median split and subdivided participants in each of the 2 age groups into high and low performing groups (for similar analyses with performance-based groupings see Chowdhury et al, 2013;Nagel et al, 2009). Although the use of median split analyses has been discouraged previously because of loss of power and information as well as the potential increase in type 1 errors (MacCallum et al, 2002), findings from recent simulation studies suggest these issues are only of concerns in situations in which multiple correlated predictor variables are used (Iacobucci et al, 2014), which is not the case in the present study. Furthermore, studies that criticized the use of median splits (see MacCallum et al, 2002) also acknowledged that 1 area where these procedures are potentially useful is in situations in which performance is a potential moderator of other relationships (e.g., brain-behavior relationships).…”

Section: Behavioral Data Analysismentioning

confidence: 99%

Age-related prefrontal impairments implicate deficient prediction of future reward in older adults

Eppinger

Heekeren

2015

Neurobiology of Aging

View full text Add to dashboard Cite

Foresighted decision-making depends on the ability to learn the value of future outcomes and the sequential choices necessary to achieve them. Using a 3-stage Markov decision task and functional magnetic resonance imaging, we investigated age differences in the ability to extract state transition structures while learning to predict future reward. In younger adults learning was associated with enhanced activity in the prefrontal cortex (PFC). In older adults (OA) we found no evidence for PFC recruitment. However, high-performing OA showed enhanced striatal activity, suggesting that they may engage in a model-free (experience-based) learning strategy. Change point analyses revealed that in younger adults learning was characterized by distinct and abrupt shifts in PFC activity, which were predictive of behavioral change points. In OA PFC activity was less pronounced and not predictive of behavior. Our findings suggest that age-related impairments in learning future reward value can be attributed to a deficit in extracting sequential state transition structures. This deficit may lead to myopic decisions in OA if contextual information has to be temporally integrated.

show abstract

Toward a More Nuanced Understanding of the Statistical Properties of a Median Split

Cited by 64 publications

References 34 publications

Knee extensor power asymmetry is unrelated to functional mobility of older adults

Knee extensor power asymmetry is unrelated to functional mobility of older adults

A clearer spotlight on spotlight: Understanding, conducting and reporting

Age-related prefrontal impairments implicate deficient prediction of future reward in older adults

Contact Info

Product

Resources

About