*2.1.2. Key Informant interviews*

**2. Methods**

138 Tsunami

**2.1. Primary data collection**

answers for both parties (**Figure 3**).

*2.1.1. Household interviews*

operation of these early warning communication systems.

The study methods included desk review of the available documents and data on the subject, collection and analysis of primary and secondary sources data, and direct field observations as well as interviews with community's representatives and concerned personnel. All such information was finally summarized and put for analysis to identify issues and gaps, and to develop recommendations on effective early warning to coastal communities of Pakistan.

Primary data collection is generally based on field survey comprising three main parts: (a)

The research team, led by the principal investigator, visited ten villages in Thatta, Sujawal, and Badin districts in Sindh and in the Gwadar district of Balochistan. The team was assisted by representatives of the communities for interpretation and to translate the questions and

**Figure 3.** Discussion carried out with women at Sonth Village living at the bank of Basool River where it falls into sea.

A detailed questionnaires were designed for household interviews that focused primarily on the basic infrastructure related to early warning communication systems present in the communities, their efficacy in delivering information and warning the potential affectees in times of disaster (cyclones or tsunamis), and constraints or limitations observed in the

household interviews, (b) key informant interviews, and (c) focus group discussions.

For this part, the research team interviewed senior officials from various organizations namely WWF‐Pakistan, Rural Community Development Council (RCDC) Gwadar, Trust for Conser‐ vation of Coastal Resources (TCCR) Sindh, Pakistan Meteorological Department, Pakistan Navy and the Government of Sindh. Their observations, suggestions, and recommendations pertaining to the gaps present in the currently employed early warning systems and conse‐ quently the required improvement have been incorporated in the chapter.
