Equivalence Tests

Lakens, Daniël

doi:10.1177/1948550617697177

Cited by 1,358 publications

(652 citation statements)

References 41 publications

Supporting

Mentioning

577

Contrasting

Unclassified

Order By: Relevance

“…A TOST procedure indicated that baseline reaction times were equivalent in both conditions (i.e. the observed effect size of d z = -0.02 was within the equivalent bounds of a moderate effect size of d z = -0.4 and 0.4; t(31)=2.15, p=0.02 [20]). The baseline RTs (Block 1) were then subtracted from each subsequent block to create a ΔRT value for each of Blocks 2-4 ( Figure 3B).…”

Section: Reaction Timesmentioning

confidence: 98%

The Time Course of Ineffective Sham Blinding During 1mA tDCS

Greinacher¹,

Buhôt²,

Möller³

et al. 2018

Preprint

View full text Add to dashboard Cite

Background: Studies using transcranial direct current stimulation (tDCS) typically compare the effects of an active (10-30min) relative to a shorter sham (placebo) protocol. Both active and sham tDCS are presumed to be perceptually identical on the scalp, and thus represent an effective method of delivering double-blinded experimental designs. However, participants often show above-chance accuracy when asked which condition involved active/sham retrospectively.Objective/Hypothesis: We aimed to assess the time course of sham-blinding during active and sham tDCS. We predicted that 1) Participants will be aware that the current is switched on for a longer duration in the active versus the sham protocol, 2) Active anodal tDCS will reduce reaction times more effectively than sham.Methods: 32 adults were tested in a pre-registered, double-blinded, within-subjects design. A forced-choice reaction time task was undertaken before, during and after active (10min 1mA) andsham (20s 1mA) tDCS. The anode was placed over the left primary motor cortex (C3) to target the right hand, and the cathode on the right forehead. Two probe questions were asked every 30s: "Is the stimulation on? "and "How sure are you?".Results: Distinct periods of non-overlapping confidence intervals were identified between the active and sham conditions, totalling 5min (57.1% of the total difference in stimulation time). These began immediately after sham ramp-down and lasted until the active protocol had ended. Active tDCS had no effect on reaction times compared to sham (ΔRT active vs sham p>0.38 in all blocks). Conclusions:We show a failure of placebo control during low-intensity tDCS.

show abstract

Section: Reaction Timesmentioning

confidence: 98%

The Time Course of Ineffective Sham Blinding During 1mA tDCS

Greinacher¹,

Buhôt²,

Möller³

et al. 2018

Preprint

View full text Add to dashboard Cite

show abstract

“…Two one-sided t-tests (TOSTs) evaluated equivalence between each pairwise comparison [52] as implemented in the `equivalence` package in R. TOSTs require an upper and lower bound effect size. Due to increased sample sizes required to conduct TOSTs [52], effect sizes conventionally considered to be medium magnitude were first examined, setting our equivalence boundary at d = -.5 and .5, respectively. A follow-up analysis used an effect size boundary from d = -.3 to .3 to compare Occasional and Non-Users (Frequent User comparisons were not conducted at this boundary due to limited power).…”

Section: Statistical Analysesmentioning

confidence: 99%

Cannabis Use in Youth is Associated with Limited Alterations in Brain Structure

Scott

Rosen

Moore

et al. 2018

Preprint

View full text Add to dashboard Cite

Frequent cannabis use during adolescence has been associated with alterations in brain structure. However, studies have featured relatively inconsistent results, predominantly from small samples, and few studies have examined less frequent users to shed light on potential brain structure differences across levels of cannabis use. In this study, high-resolution T1weighted MRIs were obtained from 781 youth aged 14-21 years who were studied as part of the Philadelphia Neurodevelopmental Cohort. This sample included 147 cannabis users (109 Occasional [≤1-2 times per week] and 38 Frequent [≥ 3 times per week] Users) and 634cannabis Non-Users. Several structural neuroimaging measures were examined in whole brain analyses, including gray and white matter volumes, cortical thickness, and gray matter density.Established procedures for stringent quality control were conducted, and two automated neuroimaging software processing packages were used to ensure robustness of results. There were no significant differences by cannabis group in global or regional brain volumes, cortical thickness, or gray matter density, and no significant group by age interactions were found.Follow-up analyses indicated that values of structural neuroimaging measures by cannabis group were similar across regions, and any differences among groups were likely of a small magnitude. In sum, structural brain metrics were similar among adolescent and young adult cannabis users and non-users. Our data converge with prior large-scale studies suggesting small or limited associations between cannabis use and structural brain measures in youth.Detailed studies of vulnerability to structural brain alterations and longitudinal studies examining long-term risk are indicated.Note. K-W = Kruskall-Wallis; FCU = Frequent Cannabis User; NU = Non-User; OCU = Occasional Cannabis User.

show abstract

“…In particular, participants who played the first-price auction against a computer who stopped bidding at 46cents drive this result, as they were bidding significantly more (M = 59.9, SEM = 3.7) than participants who played the first-price auction against a computer player who stopped bidding at 50 cents (M = 48.8, SEM = 4.2, t(206) = 1.99, p = .048, d = 0.27). Even in this case, we can reject the hypothesis that direct experience in a competitive situation without the escalation aspect would reduce escalation by 1 cent or more (p = .04, following the TOST procedure,Lakens, 2017).Supporting Hypothesis 4, participants with indirect experience of vicariously learning the consequences of a competitive escalation situation escalated less (M = 34.0, SEM = 2.29) in a subsequent competitive escalation situation than other participants without such indirect experience (t(393) = 2.47, p = .01, d = 0.24). Supporting Hypothesis 5, participants who mentally simulated how the competitive escalation situation would unfold and set themselves a limit did not escalate significantly less in a subsequent competitive escalation situation (M = 43.0, SEM = 2.42) than participants without such an indirect…”

mentioning

confidence: 68%

Competitive escalation and interventions

Hafenbrädl

Woike

2018

Behavioral Decision Making

View full text Add to dashboard Cite

Competitive escalation occurs frequently in managerial environments, when decisions create sunk costs and decision makers compete under time pressure. In a series of experiments using a minimal dollar auction paradigm, we test interventions to prevent competitive escalation. Without any intervention, most people, including experienced managers, escalate and lose money by bidding more than the price is worth (e.g., more than 10 € for 10 €). We test several interventions, in which we provide individuals with different types of experience: direct experience in structurally identical and in structurally similar situations, as well as direct experience in similarly competitive situations (lacking the escalation dimension). We also study indirect experience based on vicariously learning about the situation's consequences (experienced by others) and based on mental simulation by setting oneself a limit regarding where to exit the competition. In 3 experiments (N = 1,229), we find that direct experience in exactly the same or a structurally similar situation allows individuals to prevent subsequent escalation, whereas direct experience in a similar situation without escalation does not. Indirect experience based on vicarious learning successfully reduces competitive escalation, whereas a goal‐setting intervention that has proven instrumental in reducing classic escalation of commitment is not effective. This pattern of variation in the effectiveness of different interventions is consistent with the theory of a hot‐cold empathy gap that prevents people from anticipating how they will experience a competitive situation before entering it. As a methodological contribution, we developed a deception‐free computer‐player dollar‐auction for online participants and a dynamic chicken game.

show abstract

Equivalence Tests

Cited by 1,358 publications

References 41 publications

The Time Course of Ineffective Sham Blinding During 1mA tDCS

The Time Course of Ineffective Sham Blinding During 1mA tDCS

Cannabis Use in Youth is Associated with Limited Alterations in Brain Structure

Competitive escalation and interventions

Contact Info

Product

Resources

About