Tijdink J K, Smulders Y M, Vergouwen A C M, de Vet H C W, Knol D L
Department of Internal Medicine, VU University Medical Centre, PO Box 7057, 1007 MB, Amsterdam, The Netherlands,
Qual Life Res. 2014 Sep;23(7):2055-62. doi: 10.1007/s11136-014-0643-6. Epub 2014 Feb 13.
To determine content validity, structural validity, construct validity and reliability of an internet-based questionnaire designed for assessment of publication pressure experienced by medical scientists.
The Publication Pressure Questionnaire (PPQ) was designed to assess psychological pressure to publish scientific papers. Content validity was evaluated by collecting independent comments from external experts (n = 7) on the construct, comprehensiveness and relevance of the PPQ. Structural validity was assessed by factor analysis and item response theory (IRT) using the generalized partial credit model. Pearson's correlation coefficients were calculated to assess potential correlations with the emotional exhaustion and depersonalization subscales of the Maslach Burnout Inventory (MBI). Single test reliability (lambda2) was obtained from the IRT analysis.
Content validity was satisfactory. Confirmatory factor analysis did not support the presence of three initially assumed separate domains of publication pressure (i.e., personally experienced publication pressure, publication pressure in general, pressure on position of scientist). After exclusion of the third domain (six items), we performed exploratory factor analysis and IRT. The goodness-of-fit statistics for the IRT assuming a single dimension were satisfactory when four items were removed, resulting in 14 items of the final PPQ. Correlations with the emotional exhaustion and depersonalization scales of the MBI were 0.34 and 0.31, respectively, supporting construct validity. Single test administration reliability lambda2 was 0.69 and 0.90 on the test scores and expected a posteriori scores, respectively.
The PPQ seems a valid and reliable instrument to measure publication pressure among medical scientists.
确定一份用于评估医学科学家所经历的发表压力的基于互联网的问卷的内容效度、结构效度、构想效度和信度。
发表压力问卷（PPQ）旨在评估发表科学论文的心理压力。通过收集外部专家（n = 7）对PPQ的结构、全面性和相关性的独立意见来评估内容效度。使用广义部分计分模型通过因素分析和项目反应理论（IRT）评估结构效度。计算Pearson相关系数以评估与马氏倦怠量表（MBI）的情感耗竭和去个性化子量表的潜在相关性。从IRT分析中获得单测信度（lambda2）。
内容效度令人满意。验证性因素分析不支持最初假设的发表压力的三个单独领域（即个人经历的发表压力、一般发表压力、科学家职位压力）的存在。在排除第三个领域（六个项目）后，我们进行了探索性因素分析和IRT。当去除四个项目时，假设为单维度的IRT的拟合优度统计量令人满意，最终PPQ产生14个项目。与MBI的情感耗竭和去个性化量表的相关性分别为0.34和0.31，支持构想效度。单测施测信度lambda2在测试分数和预期后验分数上分别为0.69和0.90。
PPQ似乎是一种有效且可靠的工具，可用于测量医学科学家的发表压力。
