Sociological practitioners often face considerable methodological uncertainty when undertaking a quantitative analysis. This methodological uncertainty encompasses both data construction (e.g. defining variables) and analysis (e.g. selecting and specifying a modelling procedure). Methodological uncertainty can lead to results that are fragile and arbitrary. Yet, many practitioners may be unaware of the potential scale of methodological uncertainty in quantitative analysis, and the recent emergence of techniques for addressing it. Recent proposals for ‘multi-strategy’ approaches seek to identify and manage methodological uncertainty in quantitative analysis. We present a case-study of a multi-strategy analysis, applied to the problem of estimating the long-term impact of 1980s UK government-sponsored youth training. We use this case study to further highlight the problem of cumulative methodological fragilities in applied quantitative sociology and to discuss and help develop multi-strategy analysis as a tool to address them.

社会学从业人员在进行定量分析时通常面临相当大的方法论不确定性。这种方法上的不确定性包括数据构造（例如，定义变量）和分析（例如，选择和指定建模程序）。方法学的不确定性可能导致结果脆弱且武断。然而，许多从业者可能尚未意识到定量分析中方法学不确定性的潜在规模以及用于解决该问题的技术的最新出现。最近关于“多策略”方法的建议试图识别和管理定量分析中的方法学不确定性。我们提供了一项多策略分析的案例研究，用于评估1980年代英国政府资助的青年培训的长期影响的问题。