*2.2. Data Collection*

In each district, two maize producing villages were randomly selected. About 33 maize farmers were face-to-face interviewed per village by surveyors using a standardized questionnaire. The final sample consisted of 1237 maize farmers. The surveys were conducted from October to December 2018. The data collected included the socio-economic profile of farmers (Table 2). In addition, data were collected on farm characteristics, knowledge, and perceptions of *S. frugiperda*, periods of attack, severity of damage, yield losses, trends in the spread of *S. frugiperda* attacks, farmer managemen<sup>t</sup> practices, and their effectiveness.


**Table 2.** Descriptive statistics on the socio-economic profiles of the surveyed farmers.

Yield losses in this study were estimated by farmers. They compared maize yield before and after the invasion of *S. frugiperda*. To prevent farmers from confusing attacks of *S. frugiperda* with those of other maize pests, colour photographic images showing the condition of a field attacked by *S. frugiperda*, the stages of *S. frugiperda* development, the severity of attack and other maize pests frequently encountered in Benin were included in the questionnaire.

The Likert scale techniques [30] have been used to collect data on the managemen<sup>t</sup> practices effectiveness among farmers who have applied at least one managemen<sup>t</sup> practice. The Likert scale used has been labelled as follows: 1—totally ine ffective, 2—ine ffective, 3—relatively ine ffective, 4—indi fferent, 5—relatively e ffective, 6—e ffective, and 7—totally e ffective.
