Table 1 Limitations of reference class... | Emerald Publishing

Table 1

Limitations of reference class forecasting

Reference	Method	Content	Industry	Limitations
Zani et al. (2024)	RCF	Proposing an alternative method for class selection before RCF application	Infrastructure	Lack of large samples of similar projects; difficulty in gathering enough data with the accurate information necessary to form the reference classes; the selection of reference projects is a biased process; the method becomes the worst-performing contingency estimating method when the reference class is not specific enough
Zani and Adey (2025)	RCF + alternative stratified approach	Employs RCF for Swiss Highway project cost forecasting	Construction industry	RCF generates a subjective single uplift value rather than providing a spectrum of uplifts that reflect varying degrees of certainty
Themsen (2019)	RCF	Case study on the application of RCF in an infrastructure project	Infrastructure	The application of RCF did not prevent the experts from applying their own biased judgment when selecting the reference class of projects
Salling and Leleur (2015)	RSF	Propose the use of RSF – Integration of RCF and Quantitative Risk Analysis	Transport	RCF initial input is often wrong and biased. Try to solve this by integrating Monte Carlo Simulation and Risk Analysis
Bayram and Al-Jibouri (2016)	RCF	Application RCF to construction projects cost estimate in Turkey	Construction industry	RCF relies on a single uplift value. This approach does not account for varying levels of risk that different projects might entail. The paper proposes an improvement by integrating a range of uplift values that correspond to different risk levels
Leleur et al. (2015)	SIMRISK (RCF + OT + EJ)	Apply reference class forecasting (RCF) in association with risk simulation tools	Infrastructure	RCF can be effectively applied but must be used in a flexible way with the other tools to cope with possible mistakes in the sample selection
Kaiser and Snyder (2012)	RCF + Regression model	Offshore wind capital cost estimation	Energy infrastructure	Inconsistent reporting standards and varying detail levels in data sources can introduce biases in the forecasting; fluctuating exchange rates and specific inflation rates introduce; conversion errors and biases; Variabilities in project conditions necessitate normalization, limiting the effectiveness of cost comparisons
Lovallo et al. (2012)	RCF and Similarity Based Forecasting (SBF)	Examine model of analogy and using empirical test compare it with RCF	Strategic management	Limits related to subjective expected utility (selection bias and anchoring effect)

Reference	Method	Content	Industry	Limitations
Zani et al. (2024)	RCF	Proposing an alternative method for class selection before RCF application	Infrastructure	Lack of large samples of similar projects; difficulty in gathering enough data with the accurate information necessary to form the reference classes; the selection of reference projects is a biased process; the method becomes the worst-performing contingency estimating method when the reference class is not specific enough
Zani and Adey (2025)	RCF + alternative stratified approach	Employs RCF for Swiss Highway project cost forecasting	Construction industry	RCF generates a subjective single uplift value rather than providing a spectrum of uplifts that reflect varying degrees of certainty
Themsen (2019)	RCF	Case study on the application of RCF in an infrastructure project	Infrastructure	The application of RCF did not prevent the experts from applying their own biased judgment when selecting the reference class of projects
Salling and Leleur (2015)	RSF	Propose the use of RSF – Integration of RCF and Quantitative Risk Analysis	Transport	RCF initial input is often wrong and biased. Try to solve this by integrating Monte Carlo Simulation and Risk Analysis
Bayram and Al-Jibouri (2016)	RCF	Application RCF to construction projects cost estimate in Turkey	Construction industry	RCF relies on a single uplift value. This approach does not account for varying levels of risk that different projects might entail. The paper proposes an improvement by integrating a range of uplift values that correspond to different risk levels
Leleur et al. (2015)	SIMRISK (RCF + OT + EJ)	Apply reference class forecasting (RCF) in association with risk simulation tools	Infrastructure	RCF can be effectively applied but must be used in a flexible way with the other tools to cope with possible mistakes in the sample selection
Kaiser and Snyder (2012)	RCF + Regression model	Offshore wind capital cost estimation	Energy infrastructure	Inconsistent reporting standards and varying detail levels in data sources can introduce biases in the forecasting; fluctuating exchange rates and specific inflation rates introduce; conversion errors and biases; Variabilities in project conditions necessitate normalization, limiting the effectiveness of cost comparisons
Lovallo et al. (2012)	RCF and Similarity Based Forecasting (SBF)	Examine model of analogy and using empirical test compare it with RCF	Strategic management	Limits related to subjective expected utility (selection bias and anchoring effect)

Source(s): Authors’ own creation