Abstract
A consumer acceptance study was undertaken to determine attributes influencing acceptance and purchase intent of fresh water fish products developed by MARDI. A number of statistical techniques such as analysis of variance (ANOVA), multivariate analysis of variance (MANOVA), descriptive discriminant analysis (DDA) and logistic regression analysis (LRA) were employed to determine crucial underlying attributes that influence preference and
purchase intent. ANOVA and MANOVA were employed to find out whether there was significant difference between products reviewed. ANOVA considered each attribute individually while MANOVA considered all attributes simultaneously. It was found that all products were significantly different. Subsequently, DDA was performed to identify which attributes contributed to the difference. Attributes taste and overall acceptability were responsible for the difference in the consumer preference. DDA indicated that taste with pooled within canonical structure of 0.904796 was the important attribute, followed by overall acceptability with pooled within canonical structure of 0.897147 in determining customer preference towards fresh water fish products developed by MARDI.

Abstrak
Penerimaan pengguna terhadap produk ikan air tawar yang dibangunkan oleh MARDI telah dikaji. Objektif utama adalah untuk mengenal pasti atribut yang penting dan diutamakan oleh pengguna dalam menentukan penerimaan dan niat untuk membeli produk yang dikaji. Beberapa teknik analisis statistik iaitu ‘analysis of variance’ (ANOVA), ‘multivariate analysis of variance’ (MANOVA), ‘descriptive discriminant analysis’ (DDA) and ‘logistic regression analysis’ (LRA) telah digunakan bagi tujuan tersebut. Teknik ANOVA dan MANOVA digunakan untuk menentukan sama ada terdapat perbezaan ketara antara produk daripada sudut penerimaan pengguna terhadap produk tersebut. ANOVA mengambil kira jika terdapat perbezaan ketara apabila tiap-tiap atribut dikendalikan berasingan, sementara MANOVA mengambil kira perbezaan apabila semua atribut dikendalikan serentak. Kedua-dua analisis statistik tersebut menunjukkan terdapat perbezaan ketara antara produk daripada sudut penerimaan pengguna. Seterusnya analisis DDA dilakukan untuk menentukan atribut yang menyebabkan perbezaan ketara tersebut. Didapati bahawa atribut rasa dengan nilai statistik struktur ‘pooled within canonical’, 0.904796 dan penerimaan keseluruhan, 0.897147 ialah dua atribut yang amat diutamakan oleh pengguna semasa menentukan penerimaan dan niat untuk membeli produk yang dikaji.