On probability-based inference under data missing by design

Whether a statistician wants to complement a probability model for observed data with a prior distribution and carry out fully probabilistic inference, or base the inference only on the likelihood function, may be a fundamental question in theory, but in practice it may well be of less importance if the likelihood contains much more information than the prior. Maximum likelihood inference can be justified as a Gaussian approximation at the posterior mode, using flat priors. However, in situations where parametric assumptions in standard statistical models would be too rigid, more flexible model formulation, combined with fully probabilistic inference, can be achieved using hierarchical Bayesian parametrization. This work includes five articles, all of which apply probability modeling under various problems involving incomplete observation. Three of the papers apply maximum likelihood estimation and two of them hierarchical Bayesian modeling. Because maximum likelihood may be presented as a special case of Bayesian inference, but not the other way round, in the introductory part of this work we present a framework for probability-based inference using only Bayesian concepts. We also re-derive some results presented in the original articles using the toolbox equipped herein, to show that they are also justifiable under this more general framework. Here the assumption of exchangeability and de Finetti's representation theorem are applied repeatedly for justifying the use of standard parametric probability models with conditionally independent likelihood contributions. It is argued that this same reasoning can be applied also under sampling from a finite population. The main emphasis here is in probability-based inference under incomplete observation due to study design. This is illustrated using a generic two-phase cohort sampling design as an example. The alternative approaches presented for analysis of such a design are full likelihood, which utilizes all observed information, and conditional likelihood, which is restricted to a completely observed set, conditioning on the rule that generated that set. Conditional likelihood inference is also applied for a joint analysis of prevalence and incidence data, a situation subject to both left censoring and left truncation. Other topics covered are model uncertainty and causal inference using posterior predictive distributions. We formulate a non-parametric monotonic regression model for one or more covariates and a Bayesian estimation procedure, and apply the model in the context of optimal sequential treatment regimes, demonstrating that inference based on posterior predictive distributions is feasible also in this case.

Uncertainty in forest simulators and forest planning systems

The forest simulator is a computerized model for predicting forest growth and future development as well as effects of forest harvests and treatments. The forest planning system is a decision support tool, usually including a forest simulator and an optimisation model, for finding the optimal forest management actions. The information produced by forest simulators and forest planning systems is used for various analytical purposes and in support of decision making. However, the quality and reliability of this information can often be questioned. Natural variation in forest growth and estimation errors in forest inventory, among other things, cause uncertainty in predictions of forest growth and development. This uncertainty stemming from different sources has various undesirable effects. In many cases outcomes of decisions based on uncertain information are something else than desired. The objective of this thesis was to study various sources of uncertainty and their effects in forest simulators and forest planning systems. The study focused on three notable sources of uncertainty: errors in forest growth predictions, errors in forest inventory data, and stochastic fluctuation of timber assortment prices. Effects of uncertainty were studied using two types of forest growth models, individual tree-level models and stand-level models, and with various error simulation methods. New method for simulating more realistic forest inventory errors was introduced and tested. Also, three notable sources of uncertainty were combined and their joint effects on stand-level net present value estimates were simulated. According to the results, the various sources of uncertainty can have distinct effects in different forest growth simulators. The new forest inventory error simulation method proved to produce more realistic errors. The analysis on the joint effects of various sources of uncertainty provided interesting knowledge about uncertainty in forest simulators.

Valuation of reduced eutrophication in the Gulf of Finland : Choice experiment with attention to heterogeneous and discontinuous preferences and respondent uncertainty

Eutrophication of the Baltic Sea is a serious problem. This thesis estimates the benefit to Finns from reduced eutrophication in the Gulf of Finland, the most eutrophied part of the Baltic Sea, by applying the choice experiment method, which belongs to the family of stated preference methods. Because stated preference methods have been subject to criticism, e.g., due to their hypothetical survey context, this thesis contributes to the discussion by studying two anomalies that may lead to biased welfare estimates: respondent uncertainty and preference discontinuity. The former refers to the difficulty of stating one s preferences for an environmental good in a hypothetical context. The latter implies a departure from the continuity assumption of conventional consumer theory, which forms the basis for the method and the analysis. In the three essays of the thesis, discrete choice data are analyzed with the multinomial logit and mixed logit models. On average, Finns are willing to contribute to the water quality improvement. The probability for willingness increases with residential or recreational contact with the gulf, higher than average income, younger than average age, and the absence of dependent children in the household. On average, for Finns the relatively most important characteristic of water quality is water clarity followed by the desire for fewer occurrences of blue-green algae. For future nutrient reduction scenarios, the annual mean household willingness to pay estimates range from 271 to 448 and the aggregate welfare estimates for Finns range from 28 billion to 54 billion euros, depending on the model and the intensity of the reduction. Out of the respondents (N=726), 72.1% state in a follow-up question that they are either Certain or Quite certain about their answer when choosing the preferred alternative in the experiment. Based on the analysis of other follow-up questions and another sample (N=307), 10.4% of the respondents are identified as potentially having discontinuous preferences. In relation to both anomalies, the respondent- and questionnaire-specific variables are found among the underlying causes and a departure from standard analysis may improve the model fit and the efficiency of estimates, depending on the chosen modeling approach. The introduction of uncertainty about the future state of the Gulf increases the acceptance of the valuation scenario which may indicate an increased credibility of a proposed scenario. In conclusion, modeling preference heterogeneity is an essential part of the analysis of discrete choice data. The results regarding uncertainty in stating one s preferences and non-standard choice behavior are promising: accounting for these anomalies in the analysis may improve the precision of the estimates of benefit from reduced eutrophication in the Gulf of Finland.

Digital elevation model error in terrain analysis

Digital elevation models (DEMs) have been an important topic in geography and surveying sciences for decades due to their geomorphological importance as the reference surface for gravita-tion-driven material flow, as well as the wide range of uses and applications. When DEM is used in terrain analysis, for example in automatic drainage basin delineation, errors of the model collect in the analysis results. Investigation of this phenomenon is known as error propagation analysis, which has a direct influence on the decision-making process based on interpretations and applications of terrain analysis. Additionally, it may have an indirect influence on data acquisition and the DEM generation. The focus of the thesis was on the fine toposcale DEMs, which are typically represented in a 5-50m grid and used in the application scale 1:10 000-1:50 000. The thesis presents a three-step framework for investigating error propagation in DEM-based terrain analysis. The framework includes methods for visualising the morphological gross errors of DEMs, exploring the statistical and spatial characteristics of the DEM error, making analytical and simulation-based error propagation analysis and interpreting the error propagation analysis results. The DEM error model was built using geostatistical methods. The results show that appropriate and exhaustive reporting of various aspects of fine toposcale DEM error is a complex task. This is due to the high number of outliers in the error distribution and morphological gross errors, which are detectable with presented visualisation methods. In ad-dition, the use of global characterisation of DEM error is a gross generalisation of reality due to the small extent of the areas in which the decision of stationarity is not violated. This was shown using exhaustive high-quality reference DEM based on airborne laser scanning and local semivariogram analysis. The error propagation analysis revealed that, as expected, an increase in the DEM vertical error will increase the error in surface derivatives. However, contrary to expectations, the spatial au-tocorrelation of the model appears to have varying effects on the error propagation analysis depend-ing on the application. The use of a spatially uncorrelated DEM error model has been considered as a 'worst-case scenario', but this opinion is now challenged because none of the DEM derivatives investigated in the study had maximum variation with spatially uncorrelated random error. Sig-nificant performance improvement was achieved in simulation-based error propagation analysis by applying process convolution in generating realisations of the DEM error model. In addition, typology of uncertainty in drainage basin delineations is presented.

Environmental factors and uncertainty in fisheries management in the northern Baltic Sea

The Baltic Sea is a geologically young, large brackish water basin, and few of the species living there have fully adapted to its special conditions. Many of the species live on the edge of their distribution range in terms of one or more environmental variables such as salinity or temperature. Environmental fluctuations are know to cause fluctuations in populations abundance, and this effect is especially strong near the edges of the distribution range, where even small changes in an environmental variable can be critical to the success of a species. This thesis examines which environmental factors are the most important in relation to the success of various commercially exploited fish species in the northern Baltic Sea. It also examines the uncertainties related to fish stocks current and potential status as well as to their relationship with their environment. The aim is to quantify the uncertainties related to fisheries and environmental management, to find potential management strategies that can be used to reduce uncertainty in management results and to develop methodology related to uncertainty estimation in natural resources management. Bayesian statistical methods are utilized due to their ability to treat uncertainty explicitly in all parts of the statistical model. The results show that uncertainty about important parameters of even the most intensively studied fish species such as salmon (Salmo salar L.) and Baltic herring (Clupea harengus membras L.) is large. On the other hand, management approaches that reduce uncertainty can be found. These include utilising information about ecological similarity of fish stocks and species, and using management variables that are directly related to stock parameters that can be measured easily and without extrapolations or assumptions.

Real Estate Investment and Uncertainty: Econometric Modelling using Finnish Data

The purpose of this paper is to test for the effect of uncertainty in a model of real estate investment in Finland during the hihhly cyclical period of 1975 to 1998. We use two alternative measures of uncertainty. The first measure is the volatility of stock market returns and the second measure is the heterogeneity in the answers of the quarterly business survey of the Confederation of Finnish Industry and Employers. The econometric analysis is based on the autoregressive distributed lag (ADL) model and the paper applies a 'general-to-specific' modelling approach. We find that the measure of heterogeneity is significant in the model, but the volatility of stock market returns is not. The empirical results give some evidence of an uncertainty-induced threshold slowing down real estate investment in Finland.