A new integrated framework for training needs analysis

29 T raining and HR professionals consistently emphasise the importance of linking training to relevant results, as is seen by the popularity of literature on both resultsoriented training (eg Robinson and Robinson, 1989; Rummler and Brache, 1990) and training evaluation (eg Brinkerhoff, 1987; Carnevale, 1990; Jackson, 1989; Phillips, 1994). Linking training with results ± for example, improved productivity, quality and service ± is critical for justifying to management the cost of training and for maintaining the training function, especially during lean times when accountability for expenditure on training is a high priority. Decisions about whether or not to provide training for employees, and what type of training should be provided, have typically been based on the determination of training needs within an organisation. For over 30 years, however, training theorists and researchers have bemoaned the reliance by organisations on ad hoc training decisions and the lack of application of more systematic approaches to training needs analysis (TNA). Even in recent times, surveys of how TNA is actually conducted illustrate that ad hoc processes are still the norm (Holden, 1991; O’Driscoll and Taylor, 1992; Saari et al, 1988), and there continues to be a gap between recommendations from researchers and what is actually practised. Latham (1988) has suggested that this gap is due to theory and research failing to ® lter through to practitioners. In addition to this, in this article we argue that bridging the research-practice gap could be facilitated by improving the existing conceptual framework which guides training needs analysis and training decisions. Most treatments of training design and implementation refer to three interdependent phases ± needs analysis, delivery and evaluation. Needs analysis is the most crucial time for establishing links between training and results because initial decisions are made concerning what training will be provided in organisations. TNA includes the establishment of training objectives, and influences how training will be developed, delivered and evaluated (Goldstein, 1993). Despite the importance of TNA to the entire training process, there has been little theoretical development (Goldstein, 1991; Moore and Dutton, 1978; Ostroff and Ford, 1989), particularly in terms of linking training initiatives to organisationally-valued results. We propose a conceptual model which combines training with results in the needs analysis phase of training design by integrating critical aspects of two traditional TNA frameworks. After reviewing existing TNA theory, we describe our model and use it to organise a focused review of the training literature particularly relevant to TNA. We then distinguish two different focuses of training: results-focused training, which is aimed at enhancing organisationally-valued outcomes by improving how employees perform existing tasks; and task-focused training, used for preparing employees to perform tasks or jobs which are new to them. We specify the inferences that must be made in the needs analysis phase for each of these two different training aims, as well as how training objectives and evaluation strategies differ. Finally, we suggest alternative informationgathering strategies for both results and task-focused TNA. A new integrated framework for training needs analysis