If the survey was designed to be weighted, it is in fact "wrong" not to use the weights. Often, there will not be much difference between the weighted and unweighted results, so not much harm would be done by failing to weight. However, in certain circumstances (e.g., a disproportionately stratified sample, where some strata are greatly over-sampled or under-sampled), failing to weight could severely distort the results.

Thanks Burke D. Grandjean for the answer. May I ask you one more question? If I decided to use the weight, should I multiply the "weight" column value for the all the values in the row correct?