In order to investigate the impact of the economic crisis on food security in Greek households, research was undertaken on two axes. The first axis consisted of qualitative research using in-depth interviews. This was carried out with the participation of the heads of 44 households (those in charge of food decisions, in both regions): 27 were conducted in the region of Central Macedonia and 17 in the region of Eastern Macedonia and Thrace. The sample was selected using the convenience sampling method, since random sampling is not a prerequisite in qualitative research (Nickel et al. 1995). However, care was taken to gather qualitative data from different cities and villages in each region, representing rural and urban areas, as well as the different socioeconomic and demographic characteristics of participants. In all interviews the same simple questionnaire schedule was used, where discussions were recorded, transcribed, and cross-sectional content analyzed. The words and phrases derived were used to develop the quantitative questionnaire.
Quantitative research was carried out during a six-month period, from the middle of January 2015 to the middle ofJuly 2015, mainly with aided self-administrated questionnaires and, in some situations, with personal interviews. Data were collected from the two regions where qualitative research took place: the region of Central Macedonia with a population of
1,880,058 residents (7 prefectures) and the region of Eastern Macedonia and Thrace with 608,182 residents (5 prefectures), according to the census of 2011 (ypes.gr). The multistage stratified sampling method was used for data collection (Kamenidou 1999). Regarding the strata used, the first stratum was the region, the second was each prefecture, and the third consisted of the large cities and large towns (urban-rural areas).
From the last stratum, data were drawn via the random sampling method, and the maps of the cities and towns (where applicable) with their building blocks were used as a sampling frame. Additionally, the head of the family, the person who was in charge of food decisions, was used as the sampling unit (Kamenidou 1999). Due to economic and time constraints, in some rural and semi-urban areas convenience sampling was also applied in this third stratum in the sampling procedure. A total of 1528 questionnaires were collected using this method, from which 1305 were valid, with a valid response rate of 85.4 %.