如何在连接语句期间避免SQL Server 2008 R2中的默认日期1900-01-01

Question

SELECT IsNull(cmg.[Advertiser],
    con.[Advertiser]) as [Advertiser] ,
    IsNull(cmg.[AdvertiserID],
    con.[AdvertiserID]) as [AdvertiserID] ,
    con.[FloodlightConfiguration] ,
    IsNull(cmg.[Campaign],
    con.[Campaign]) as [Campaign] ,
    IsNull(cmg.[CampaignID],
    con.[CampaignID]) as [CampaignID] ,
    IsNull(cmg.[Placement],
    con.[Placement]) as [Placement] ,
    IsNull(cmg.[PlacementID],con.[PlacementID]) as [PlacementID] ,
    IsNull(cmg.[Creative],con.[Creative]) as [Creative] ,
    IsNull(cmg.[CreativeID],con.[CreativeID]) as [CreativeID] ,
    IsNull(cmg.[AD],con.[AD]) as [AD] ,
    IsNull(cmg.[ADID],
    con.[ADID]) as [ADID] ,
    IsNull(cmg.[Country],
    con.[Country]) as [Country] ,
    con.[Activity] ,
    con.[ActivityID] ,
    IsNull(cmg.[DeliveryDate],0) as [ConversionDate] ,
    IsNull(con.[TotalConversions],0) as [TotalConversions] ,
    IsNull(con.[TotalRevenue],0) as [TotalRevenue] ,
    IsNull([ClickthroughConversions],0) as [ClickthroughConversions] ,
    IsNull([ClickthroughRevenue],0) as [ClickthroughRevenue] ,
    IsNull([ViewthroughConversions],0) as [ViewthroughConversions] ,
    IsNull([ViewthroughRevenue],0) as [ViewthroughRevenue] ,
    case when con.[ActivityID] is not null then 0 else cmg.Impressions end as Impressions,
    case when con.[ActivityID] IS NOT Null then 0 else cmg.Clicks end AS Clicks,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.MediaCost end as MediaCost,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.ActivityPerClick end as ActivityPerClick,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.ActivityPerThousandImpressions end as ActivityPerThousandImpressions,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.CostPerActivity end as CostPerActivity,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.CostPerClick end as CostPerClick,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.CostPerRevenue end as CostPerRevenue,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.EffectiveCPM end as EffectiveCPM ,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.RevenuePerClick end as RevenuePerClick,
    case when con.[ActivityID] IS NOT NULL then 0 else cmg.RevenuePerThousandImpressions end as RevenuePerThousandImpressions 
FROM [DM_886_MarriottSocialData].[dbo].[DFA_BSE_DIS_MD_1] cmg
    FULL OUTER JOIN [DM_886_MarriottSocialData].[dbo].[DFA_BSE_DIS_CON_1] con
        ON cmg.PlacementID = con.PlacementID 
            AND cmg.DeliveryDate = con.ConversionDate 
            AND cmg.CampaignID = con.CampaignID 
            AND cmg.Country = con.Country 
            AND cmg.ADID = con.ADID 
            AND cmg.CreativeID = con.CreativeID 
            AND cmg.AdvertiserID = con.AdvertiserID 
WHERE cmg.DeliveryDate >= '2014-01-01' 
    AND cmg.DeliveryDate <= '2014-03-16'

这是我目前用于评估聚合表的代码，但我在某些记录中获得了1900-01-01的默认日期。但是，我希望决赛桌上的所有记录。原因是由于1900-01-01，一些列的影响在各自的值中有5％-10％的波动。

由于