Dealing with baseline differences: two principles and two dilemmas

Jamieson, John

doi:10.1016/s0167-8760(98)00048-8

Cited by 41 publications

(35 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is not clear how to obtain an unbiased measure of change in such circumstances. Although there are still unanswered questions about how to measure change in the 2-phase design, the two guidelines suggested by Jamieson (1999) are a good starting point: Avoid ANCOVA except for randomized experiments, and avoid using gain scores with skewed data.…”

Section: Discussionmentioning

confidence: 99%

“…By failing to provide clear guidelines about when ANCOVA should be used, these two articles (as well as others) have left this issue open, so researchers have the option of choosing either ANCOVA or gain scores. Jamieson (1999) recently pointed out that this option creates an ethical dilemma for researchers because the two methods are too differentially powerful to detect differences in changes. Simply by looking at the direction of pretest differences and knowing which group is expected to change more, it is possible to identify which of ANCOVA or gain scores will be more powerful to detect this hypothesized difference (see Figure 4 in Jamieson, 1999, for an illustration of the patterns of differences that each of ANCOVA or gain scores are better able to detect).…”

mentioning

confidence: 99%

“…These findings are extensions of the ANCOVA bias: When the third variable is highly correlated with pretest (which is comparable to large baseline differences in ANCOVA), regression measures are biased to detect a correlation of change with the third variable that is in the same direction as the correlation with baseline (which is comparable to detecting mean differences that diverge from pretest to posttest in ANCOVA). The ethical dilemma described in Jamieson (1999) also applies to the case of a continuous variable. For example, suppose one predicts that physically fit individuals will show lower heart rate increases to stress than physically unfit.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Structural Equation Models and the Regression Bias for Measuring Correlates of Change

Cribbie

Jamieson

2000

Educational and Psychological Measurement

Self Cite

View full text Add to dashboard Cite

ANCOVA and regression both exhibit a directional bias when measuring correlates of change. This bias confounds the comparison of changes between naturally occurring groups with large pretest differences (ANCOVA), or for identifying predictors of change when the predictor is correlated with pretest (regression). This bias is described in some detail. A computer simulation study is presented, which shows that properly identified structural equation models are not susceptible to this bias. Neither gain scores (posttest minus pretest) nor structural equation models exhibit the “regression bias.” Other factors, such as skewness, that may confound measurement of change are also discussed.

show abstract

Section: Discussionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Structural Equation Models and the Regression Bias for Measuring Correlates of Change

Cribbie

Jamieson

2000

Educational and Psychological Measurement

Self Cite

View full text Add to dashboard Cite

show abstract

“…A baseline measure is recommended when problems with skewed physiological data may arise (see Jamieson, 1999). Importantly, when baseline HRV was not controlled, the effects reported remained significant.…”

Section: Data Analysis Strategymentioning

confidence: 99%

“…Work control refers to control available in the work environment and employee discretion over methods and pacing of work. Research has focused on the chronic effects of high demands and low work control on physiological health (particularly CVDs) and psychological health (e.g., Chandola et al, 2006;Peter and Siegrist, 2000), with metaanalytical research on the JD-CM showing modest support for these relationships (de Lange et al, 2003;Häusser et al, 2010;van der Doef and Maes, 1998;1999). Limited research has examined whether providing high task control facilitates better adaptation in the moment.…”

Section: Introductionmentioning

confidence: 99%

Regulatory focus moderates the relationship between task control and physiological and psychological markers of stress: A work simulation study

Parker¹,

Laurie²,

Newton³

et al. 2014

International Journal of Psychophysiology

View full text Add to dashboard Cite

The influence of snuff and smoking on bone accretion in late adolescence. The Tromsø study, Fit Futures

et al. 2021

View full text Add to dashboard Cite

Summary Areal bone mineral density (aBMD) predicts future fracture risk. This study explores associations between use of tobacco and bone accretion in Norwegian adolescents. Our results indicate that use of snuff is negatively associated with accretion of aBMD in adolescence and may be a signal of increased future fracture risk. Purpose Bone mineral accrual in childhood and adolescence is a long-term primary preventive strategy of osteoporosis. Areal bone mineral density (aBMD) is a surrogate measure of bone strength and a predictor of fracture risk. The aim of this population-based 2-year follow-up cohort study was to explore associations between use of snuff and smoking and changes (∆) in aBMD in Norwegian girls and boys aged 15–17 years at baseline. Methods The first wave of the Tromsø study, Fit Futures was conducted from 2010 to 2011. Femoral neck (FN), total hip (TH), and total body (TB) bone mineral content (BMC) and aBMD were measured by dual-energy X-ray absorptiometry. Information on use of snuff, smoking habits, and other lifestyle related variables were collected through self-administered questionnaires. Two years later, during 2012–2013, the measurements were repeated in the second wave. The present study included 349 girls and 281 boys and compared “non-users” (n = 243 girls, 184 boys) with “users” (n = 105 girls, 96 boys) of snuff and “non-smokers” (n = 327 girls, 249 boys) with “smokers” (n = 21 girls, 31 boys) using linear regression adjusted for age, baseline height and weight, change in height and weight, pubertal maturation, physical activity, ethnicity, alcohol consumption, diagnosis known to affect bone, and medication known to affect bone. The influence of “double use” on bone accretion was also explored. Results In girls, no associations between use of snuff and ∆aBMD were found. In boys, use of snuff was associated with reduced bone accretion in all ∆aBMD models. Sensitivity analysis with exclusion of “sometimes” users of snuff strengthened associations at femoral sites in girls and attenuated all associations in boys. In girls, no associations between smoking and ∆aBMD were found. In boys, only the association with TB ∆aBMD was significant in the fully adjusted models. In girls, “double users” analyses showed similar association to smoking. In boys, nearly all models showed statistically significant associations with a difference of ~ 1–2% in ∆aBMD between “non-users” and “double users” during 2 years of follow-up. Conclusions Our results indicate that tobacco use in late adolescence could be detrimental to bone accretion and may be a signal of increased fracture risk in adult life.

show abstract

Dealing with baseline differences: two principles and two dilemmas

Cited by 41 publications

References 13 publications

Structural Equation Models and the Regression Bias for Measuring Correlates of Change

Structural Equation Models and the Regression Bias for Measuring Correlates of Change

Regulatory focus moderates the relationship between task control and physiological and psychological markers of stress: A work simulation study

The influence of snuff and smoking on bone accretion in late adolescence. The Tromsø study, Fit Futures

Contact Info

Product

Resources

About